In this example, we’ll create two additional models made with CimpleG. One using only hypermethylated signatures, and the other using 3 CpGs per signature instead of just one.
+
set.seed ( 42 ) cimpleg_hyper <- CimpleG ( "fibroblasts"
)
) #> Training for target 'neurons' with 'CimpleG' has finished.: 0.407 sec elapsed #> Training for target 'glia' with 'CimpleG' has finished.: 0.55 sec elapsed #> Training for target 'blood_cells' with 'CimpleG' has finished.: 0.373 sec elapsed #> Training for target 'fibroblasts' with 'CimpleG' has finished.: 0.514 sec elapsed #> Training for target 'neurons' with 'CimpleG' has finished.: 0.342 sec elapsed #> Training for target 'glia' with 'CimpleG' has finished.: 0.304 sec elapsed #> Training for target 'blood_cells' with 'CimpleG' has finished.: 0.352 sec elapsed #> Training for target 'fibroblasts' with 'CimpleG' has finished.: 0.317 sec elapsed deconv_hyper <- run_deconvolution ( cpg_obj = cimpleg_hyper ,
@@ -282,10 +314,10 @@ "fibroblasts"
)
) #> Training for target 'neurons' with 'CimpleG' has finished.: 0.611 sec elapsed #> Training for target 'glia' with 'CimpleG' has finished.: 0.614 sec elapsed #> Training for target 'blood_cells' with 'CimpleG' has finished.: 0.703 sec elapsed #> Training for target 'fibroblasts' with 'CimpleG' has finished.: 0.431 sec elapsed #> Training for target 'neurons' with 'CimpleG' has finished.: 0.471 sec elapsed #> Training for target 'glia' with 'CimpleG' has finished.: 0.39 sec elapsed #> Training for target 'blood_cells' with 'CimpleG' has finished.: 0.459 sec elapsed #> Training for target 'fibroblasts' with 'CimpleG' has finished.: 0.403 sec elapsed deconv_3sigs <- run_deconvolution ( cpg_obj = cimpleg_3sigs ,
@@ -297,7 +329,7 @@
remember this is just an example, the results themselves are meaningless!
-
+
deconv_3sigs $ prop_3sigs <- deconv_3sigs $ proportion deconv_hyper $ prop_hyper <- deconv_hyper $ proportion deconv_result $ prop_cimpleg <- deconv_result $ proportion
first we can check how the true values compare against the predicted values
-
+
scatter_plts <- CimpleG ::: deconv_pred_obs_plot ( deconv_df = dummy_deconvolution_data ,
true_values_col = "true_vals" ,
@@ -328,12 +360,12 @@ scatter_panel <- scatter_plts |> patchwork :: wrap_plots ( ncol= 1 )
print ( scatter_panel )
+
now, more interestingly, we can see in detail and rank one of the measures used to evaluate the deconvolution results
-
+
rank_plts <- CimpleG ::: deconv_ranking_plot ( deconv_df = dummy_deconvolution_data ,
true_values_col = "true_vals" ,
@@ -345,7 +377,7 @@ rank_panel <- list ( rank_plts $ perf_boxplt [[ 1 ] ] ,rank_plts $ nemenyi_plt [[ 1 ] ] ) |> patchwork :: wrap_plots ( )
print ( rank_panel )
+
CimpleG
- 0.0.5.9024
+ 0.0.5.9028
diff --git a/docs/search.json b/docs/search.json
index a233788..5f66c36 100644
--- a/docs/search.json
+++ b/docs/search.json
@@ -1 +1 @@
-[{"path":"https://costalab.github.io/CimpleG/LICENSE.html","id":null,"dir":"","previous_headings":"","what":"GNU General Public License","title":"GNU General Public License","text":"Version 3, 29 June 2007Copyright © 2007 Free Software Foundation, Inc. .","code":" Copyright (C) This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see Copyright (C) This program comes with ABSOLUTELY NO WARRANTY; for details type 'show w'. This is free software, and you are welcome to redistribute it under certain conditions; type 'show c' for details."},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"cimpleg-simple-cpg-signatures","dir":"Articles","previous_headings":"","what":"CimpleG (Simple CpG signatures)","title":"Generate signatures","text":"CimpleG tries find CpGs best classify cell-type given train dataset also enables perform cell-type deconvolution couple easy steps can use beta M values show easy generate signatures","code":""},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"installation","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Installation","title":"Generate signatures","text":"haven’t installed CimpleG, can find instructions . However simple :","code":"if(!require(\"CimpleG\")) devtools::install_github(\"costalab/CimpleG\")"},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"loading-package","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Loading package","title":"Generate signatures","text":"load CimpleG package.","code":"library(\"CimpleG\") #> -------------------------- #> CimpleG version 0.0.5.9001 #> --------------------------"},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"loading-data","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Loading data","title":"Generate signatures","text":"tutorial, use small dataset just 409 samples 1000 CpGs. also use table metadata regarding samples. dataset comes included CimpleG. can read .","code":"# load data data(train_data) data(train_targets)"},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"running-cimpleg","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Running CimpleG","title":"Generate signatures","text":"Running CimpleG can quite simple. just need run CimpleG function parameters. generating signatures find leukocytes hepatocytes.","code":"# run CimpleG cimpleg_result <- CimpleG( train_data, train_targets, target_columns = c(\"blood_cells\",\"hepatocytes\"), train_only = TRUE ) #> Training for target 'blood_cells' with 'CimpleG' has finished.: 2.655 sec elapsed #> Training for target 'hepatocytes' with 'CimpleG' has finished.: 0.482 sec elapsed"},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"plotting-cimpleg-cpg-signature","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Plotting CimpleG CpG signature","title":"Generate signatures","text":"can quickly visualize signature able separate data.","code":"signature_plot( cimpleg_result, train_data, train_targets, sample_id_column = \"gsm\", true_label_column = \"cell_type\" ) #> $data #> # A tibble: 818 × 5 #> # Groups: sig_set [2] #> sample_id true_label signatures value sig_set #> #> 1 GSM1415516 adipocytes cg04785083 0.922 blood_cells #> 2 GSM1415516 adipocytes cg02258444 0.938 hepatocytes #> 3 GSM1415518 adipocytes cg04785083 0.932 blood_cells #> 4 GSM1415518 adipocytes cg02258444 0.912 hepatocytes #> 5 GSM1415520 adipocytes cg04785083 0.923 blood_cells #> 6 GSM1415520 adipocytes cg02258444 0.907 hepatocytes #> 7 GSM1415522 adipocytes cg04785083 0.936 blood_cells #> 8 GSM1415522 adipocytes cg02258444 0.908 hepatocytes #> 9 GSM1415526 adipocytes cg04785083 0.938 blood_cells #> 10 GSM1415526 adipocytes cg02258444 0.905 hepatocytes #> # ℹ 808 more rows #> #> $plot"},{"path":"https://costalab.github.io/CimpleG/articles/save_load_objects.html","id":"cimpleg-to-save-and-load-large-files-in-r","dir":"Articles","previous_headings":"","what":"CimpleG to save and load large files in R","title":"Quickly save and load (large) objects","text":"CimpleG comes couple neat functions save load objects using state art data compression algorithms ‘zstd’ ‘lz4’. allows much quickly save load R objects (.rds) files. Furthermore can also use functions load .rds files.","code":""},{"path":"https://costalab.github.io/CimpleG/articles/save_load_objects.html","id":"loading-the-package","dir":"Articles","previous_headings":"CimpleG to save and load large files in R","what":"Loading the package","title":"Quickly save and load (large) objects","text":"load CimpleG package.","code":"library(\"CimpleG\") #> -------------------------- #> CimpleG version 0.0.5.9003 #> --------------------------"},{"path":"https://costalab.github.io/CimpleG/articles/save_load_objects.html","id":"saving-an-object","dir":"Articles","previous_headings":"CimpleG to save and load large files in R","what":"Saving an object","title":"Quickly save and load (large) objects","text":"default use ‘zstd’ algorithm can choose different one. process pretty seamless. Let’s create large matrix save using CimpleG.","code":"my_matrix <- matrix(rnorm(100), nrow = 10, ncol = 10) save_object(my_matrix, \"my_matrix.rds\")"},{"path":"https://costalab.github.io/CimpleG/articles/save_load_objects.html","id":"loading-an-object","dir":"Articles","previous_headings":"CimpleG to save and load large files in R","what":"Loading an object","title":"Quickly save and load (large) objects","text":"load object process just seamless. need specify compression algorithm used. just work! larger file greater advantage using functions.","code":"my_loaded_matrix <- load_object(\"my_matrix.rds\") my_loaded_matrix"},{"path":"https://costalab.github.io/CimpleG/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors and Citation","text":"Tiago F.V. Maié. Author, maintainer.","code":""},{"path":"https://costalab.github.io/CimpleG/authors.html","id":"citation","dir":"","previous_headings":"","what":"Citation","title":"Authors and Citation","text":"F.V. Maié T (2023). CimpleG: CimpleG - Individual/Single CpG sites Classification Deconvolution. R package version 0.0.5.9024, https://costalab.github.io/CimpleG/, https://github.com/CostaLab/CimpleG.","code":"@Manual{, title = {CimpleG: CimpleG - Individual/Single CpG sites For Classification and Deconvolution}, author = {Tiago {F.V. Maié}}, year = {2023}, note = {R package version 0.0.5.9024, https://costalab.github.io/CimpleG/}, url = {https://github.com/CostaLab/CimpleG}, }"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"overview","dir":"","previous_headings":"","what":"Overview","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"CimpleG, R package find (small) CpG signatures.","code":""},{"path":"https://costalab.github.io/CimpleG/index.html","id":"installation","dir":"","previous_headings":"","what":"Installation","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"# Install directly from github: devtools::install_github(\"costalab/CimpleG\") # Alternatively, downloading from our release page and installing it from a local source: # - ie navigating through your system install.packages(file.choose(), repos = NULL, type = \"source\") # - ie given a path to a local source install.packages(\"~/Downloads/CimpleG_0.0.5.XXXX.tar.gz\", repos = NULL, type = \"source\") # or devtools::install_local(\"~/Downloads/CimpleG_0.0.5.XXXX.tar.gz\")"},{"path":"https://costalab.github.io/CimpleG/index.html","id":"getting-started","dir":"","previous_headings":"","what":"Getting started","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"library(\"CimpleG\") data(train_data) data(train_targets) data(test_data) data(test_targets) # check the train_targets table to see # what other columns can be used as targets # colnames(train_targets) # mini example with just 4 target signatures set.seed(42) cimpleg_result <- CimpleG( train_data = train_data, train_targets = train_targets, test_data = test_data, test_targets = test_targets, method = \"CimpleG\", target_columns = c( \"neurons\", \"glia\", \"blood_cells\", \"fibroblasts\" ) ) cimpleg_result$results # check generated signatures cimpleg_result$signatures #> neurons glia blood_cells fibroblasts #> \"cg24548498\" \"cg14501977\" \"cg04785083\" \"cg03369247\""},{"path":"https://costalab.github.io/CimpleG/index.html","id":"plot-generated-signatures","dir":"","previous_headings":"Getting started","what":"Plot generated signatures","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"# adjust target names to match signature names # check generated signatures plt <- signature_plot( cimpleg_result, train_data, train_targets, sample_id_column = \"gsm\", true_label_column = \"cell_type\" ) print(plt$plot)"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"basic-plot","dir":"","previous_headings":"Difference of means vs Sum of variances (dmsv) plots","what":"basic plot","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"plt <- diffmeans_sumvariance_plot( data = train_data, target_vector = train_targets$neurons == 1 ) print(plt)"},{"path":"https://costalab.github.io/CimpleG/index.html","id":"adding-color-highlighting-selected-features","dir":"","previous_headings":"Difference of means vs Sum of variances (dmsv) plots","what":"adding color, highlighting selected features","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"df_dmeansvar <- compute_diffmeans_sumvar( data = train_data, target_vector = train_targets$neurons == 1 ) parab_param <- .7 df_dmeansvar$is_selected <- select_features( x = df_dmeansvar$diff_means, y = df_dmeansvar$sum_variance, a = parab_param ) plt <- diffmeans_sumvariance_plot( data = df_dmeansvar, label_var1 = \"Neurons\", color_all_points = \"purple\", threshold_func = function(x, a) (a * x) ^ 2, is_feature_selected_col = \"is_selected\", func_factor = parab_param ) print(plt)"},{"path":"https://costalab.github.io/CimpleG/index.html","id":"labeling-specific-features","dir":"","previous_headings":"Difference of means vs Sum of variances (dmsv) plots","what":"labeling specific features","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"plt <- diffmeans_sumvariance_plot( data = df_dmeansvar, feats_to_highlight = cimpleg_result$signatures ) print(plt)"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"mini-example-with-just-4-signatures","dir":"","previous_headings":"Deconvolution plots","what":"mini example with just 4 signatures","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"deconv_result <- run_deconvolution( cpg_obj = cimpleg_result, new_data = test_data ) plt <- deconvolution_barplot( deconvoluted_data = deconv_result, meta_data = test_targets, sample_id = \"gsm\", true_label = \"cell_type\" ) print(plt$plot)"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"first-lets-create-additional-deconvolution-results-so-that-we-can-compare-them","dir":"","previous_headings":"Deconvolution plots > this example is a little more advanced","what":"first lets create additional deconvolution results so that we can compare them","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"example, ’ll create two additional models made CimpleG. One using hypermethylated signatures, using 3 CpGs per signature instead just one.","code":"set.seed(42) cimpleg_hyper <- CimpleG( train_data = train_data, train_targets = train_targets, test_data = test_data, test_targets = test_targets, method = \"CimpleG\", pred_type = \"hyper\", target_columns = c( \"neurons\", \"glia\", \"blood_cells\", \"fibroblasts\" ) ) #> Training for target 'neurons' with 'CimpleG' has finished.: 0.407 sec elapsed #> Training for target 'glia' with 'CimpleG' has finished.: 0.55 sec elapsed #> Training for target 'blood_cells' with 'CimpleG' has finished.: 0.373 sec elapsed #> Training for target 'fibroblasts' with 'CimpleG' has finished.: 0.514 sec elapsed deconv_hyper <- run_deconvolution( cpg_obj = cimpleg_hyper, new_data = test_data ) set.seed(42) cimpleg_3sigs <- CimpleG( train_data = train_data, train_targets = train_targets, test_data = test_data, test_targets = test_targets, method = \"CimpleG\", n_sigs = 3, target_columns = c( \"neurons\", \"glia\", \"blood_cells\", \"fibroblasts\" ) ) #> Training for target 'neurons' with 'CimpleG' has finished.: 0.611 sec elapsed #> Training for target 'glia' with 'CimpleG' has finished.: 0.614 sec elapsed #> Training for target 'blood_cells' with 'CimpleG' has finished.: 0.703 sec elapsed #> Training for target 'fibroblasts' with 'CimpleG' has finished.: 0.431 sec elapsed deconv_3sigs <- run_deconvolution( cpg_obj = cimpleg_3sigs, new_data = test_data )"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"remember-this-is-just-an-example-the-results-themselves-are-meaningless","dir":"","previous_headings":"Deconvolution plots > this example is a little more advanced","what":"remember this is just an example, the results themselves are meaningless!","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"deconv_3sigs$prop_3sigs <- deconv_3sigs$proportion deconv_hyper$prop_hyper <- deconv_hyper$proportion deconv_result$prop_cimpleg <- deconv_result$proportion dummy_deconvolution_data <- deconv_result |> dplyr::mutate(true_vals = proportion + runif(nrow(deconv_result), min=-0.1,max=0.1)) |> dplyr::select(cell_type,sample_id,prop_cimpleg,true_vals) |> dplyr::left_join(deconv_hyper |> dplyr::select(-proportion), by=c(\"sample_id\",\"cell_type\")) |> dplyr::left_join(deconv_3sigs |> dplyr::select(-proportion), by=c(\"sample_id\",\"cell_type\")) |> dplyr::mutate_if(is.numeric, function(x){ifelse(x<0,0,x)}) |> dplyr::mutate_if(is.numeric, function(x){ifelse(x>1,1,x)}) |> tibble::as_tibble()"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"first-we-can-check-how-the-true-values-compare-against-the-predicted-values","dir":"","previous_headings":"Deconvolution plots > this example is a little more advanced","what":"first we can check how the true values compare against the predicted values","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"scatter_plts <- CimpleG:::deconv_pred_obs_plot( deconv_df = dummy_deconvolution_data, true_values_col = \"true_vals\", predicted_cols = c(\"prop_cimpleg\",\"prop_hyper\",\"prop_3sigs\"), sample_id_col = \"sample_id\", group_col= \"cell_type\" ) scatter_panel <- scatter_plts |> patchwork::wrap_plots(ncol=1) print(scatter_panel)"},{"path":"https://costalab.github.io/CimpleG/index.html","id":"now-more-interestingly-we-can-see-in-detail-and-rank-one-of-the-measures-used-to-evaluate-the-deconvolution-results","dir":"","previous_headings":"Deconvolution plots > this example is a little more advanced","what":"now, more interestingly, we can see in detail and rank one of the measures used to evaluate the deconvolution results","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"rank_plts <- CimpleG:::deconv_ranking_plot( deconv_df = dummy_deconvolution_data, true_values_col = \"true_vals\", predicted_cols = c(\"prop_cimpleg\",\"prop_hyper\",\"prop_3sigs\"), sample_id_col = \"sample_id\", group_col= \"cell_type\", metrics = \"rmse\" ) rank_panel <- list(rank_plts$perf_boxplt[[1]],rank_plts$nemenyi_plt[[1]]) |> patchwork::wrap_plots() print(rank_panel)"},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG-package.html","id":null,"dir":"Reference","previous_headings":"","what":"CimpleG: CimpleG - Individual/Single CpG sites For Classification and Deconvolution — CimpleG-package","title":"CimpleG: CimpleG - Individual/Single CpG sites For Classification and Deconvolution — CimpleG-package","text":"CimpleG simple package find single-CpG based signatures use classification deconvolution.","code":""},{"path":[]},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG-package.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"CimpleG: CimpleG - Individual/Single CpG sites For Classification and Deconvolution — CimpleG-package","text":"Maintainer: Tiago F.V. Maié tiagomaie@hotmail.com","code":""},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":null,"dir":"Reference","previous_headings":"","what":"Find simple CpG (CimpleG) signatures. — CimpleG","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"Train classification model using (CpGs) features given target data.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"","code":"CimpleG( train_data, train_targets = NULL, target_columns = NULL, test_data = NULL, test_targets = NULL, method = c(\"CimpleG\", \"CimpleG_parab\", \"brute_force\", \"logistic_reg\", \"decision_tree\", \"boost_tree\", \"mlp\", \"rand_forest\"), pred_type = c(\"both\", \"hypo\", \"hyper\"), engine = c(\"glmnet\", \"xgboost\", \"nnet\", \"ranger\"), rank_method = c(\"ac_rank\", \"a_rank\", \"c_rank\"), k_folds = 10, grid_n = 10, param_p = 2, n_sigs = 1, quantile_threshold = 0.005, train_only = FALSE, split_data = FALSE, run_parallel = FALSE, deconvolution_reference = TRUE, save_dir = NULL, save_format = c(\"zstd\", \"lz4\", \"gzip\", \"bzip2\", \"xz\", \"nocomp\"), verbose = 1, targets = NULL ) cimpleg( train_data, train_targets = NULL, target_columns = NULL, test_data = NULL, test_targets = NULL, method = c(\"CimpleG\", \"CimpleG_parab\", \"brute_force\", \"logistic_reg\", \"decision_tree\", \"boost_tree\", \"mlp\", \"rand_forest\"), pred_type = c(\"both\", \"hypo\", \"hyper\"), engine = c(\"glmnet\", \"xgboost\", \"nnet\", \"ranger\"), rank_method = c(\"ac_rank\", \"a_rank\", \"c_rank\"), k_folds = 10, grid_n = 10, param_p = 2, n_sigs = 1, quantile_threshold = 0.005, train_only = FALSE, split_data = FALSE, run_parallel = FALSE, deconvolution_reference = TRUE, save_dir = NULL, save_format = c(\"zstd\", \"lz4\", \"gzip\", \"bzip2\", \"xz\", \"nocomp\"), verbose = 1, targets = NULL ) cpg( train_data, train_targets = NULL, target_columns = NULL, test_data = NULL, test_targets = NULL, method = c(\"CimpleG\", \"CimpleG_parab\", \"brute_force\", \"logistic_reg\", \"decision_tree\", \"boost_tree\", \"mlp\", \"rand_forest\"), pred_type = c(\"both\", \"hypo\", \"hyper\"), engine = c(\"glmnet\", \"xgboost\", \"nnet\", \"ranger\"), rank_method = c(\"ac_rank\", \"a_rank\", \"c_rank\"), k_folds = 10, grid_n = 10, param_p = 2, n_sigs = 1, quantile_threshold = 0.005, train_only = FALSE, split_data = FALSE, run_parallel = FALSE, deconvolution_reference = TRUE, save_dir = NULL, save_format = c(\"zstd\", \"lz4\", \"gzip\", \"bzip2\", \"xz\", \"nocomp\"), verbose = 1, targets = NULL )"},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"train_data Training dataset. matrix (s x f) methylation data (Beta values) used train/find predictors. Samples (s) must rows features/CpGs (f) must columns. train_targets data frame training target samples one-hot encoded. data frame least 1 column, many rows order `train_data`. Target columns need one-hot encoded, meaning , column target samples encoded `1` every sample encoded `0`. target_columns string specifying name column `train_targets` used training. Can character vector several columns `train_targets` used training. argument character vector, CimpleG search best predictors target sequentially parallel depending value `run_parallel` test_data Testing dataset. matrix (s x f) methylation data (Beta values) used test performance found predictors. Samples (s) must rows features/CpGs (f) must columns. `test_data` ** `test_targets` NULL, CimpleG generate stratified test dataset based `train_targets` removing 25 samples `train_data` `train_targets`. test_targets data frame testing target samples one-hot encoded. data frame least 1 column, many rows order `test_data`. Target columns need one-hot encoded, meaning , column target samples encoded `1` every sample encoded `0`. `test_data` ** `test_targets` NULL, CimpleG generate stratified test dataset based `train_targets` removing 25 samples `train_data` `train_targets`. method string specifying method type machine learning model/algorithm used training. divided two main groups. * simple models (classifiers use single feature), `CimpleG` (default), `brute_force`, `CimpleG_unscaled` `oner`; * complex models (classifiers use several features), `logistic_reg`, `decision_tree`, `boost_tree`, `mlp` `rand_forest`. pred_type string specifying type predictor/CpG searched training. used simple models. One `` (default), `hypo` `hyper`. `hypo`, hypomethylated predictors considered. `hyper`, hypermethylated predictors considered. engine string specifying machine learning engine behind `method`. used complex models. Currently use. rank_method string specifying ranking strategy rank features training. k_folds integer specifying number folds (K) used training stratified K-fold cross-validation procedure. grid_n integer specifying number hyperparameter combinations train . param_p even number `sigma / (delta^param_p)`. Tunes much weight given delta feature selection. Default 2. n_sigs Number signatures saved classification used deconvolution. Default 1. quantile_threshold number 0 1. Determines many features kept. Default 0.005. train_only boolean, TRUE, CimpleG train (find predictors) test test dataset. split_data boolean, `TRUE`, subset train data provided, creating smaller test set used test models training. parameter experimental. Default `FALSE`. run_parallel boolean, `FALSE`, default, search predictors multiple targets sequentially. `TRUE` search predictors multiple targets time (parallel processing) order save computational time. need set `future::plan()` running function. deconvolution_reference boolean, `TRUE`, create deconvolution reference matrix based training data. can later used perform deconvolution. Default `FALSE`. save_dir defined save resulting model given directory. Default NULL. save_format used save_dir NULL. One \"zstd\", \"lz4\", \"gzip\", \"bzip2\",\"xz\", \"nocomp\". zstd best option, fast compression loading times, low space usage. verbose verbose want CimpleG running. 0, message displayed, 3 every message displayed. Default 1. targets DEPRECATED use `target_columns`.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"CimpleG object results per target class.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"","code":"library(\"CimpleG\") # read data data(train_data) data(train_targets) data(test_data) data(test_targets) # run CimpleG cimpleg_result <- CimpleG( train_data = train_data, train_targets = train_targets, test_data = test_data, test_targets = test_targets, method = \"CimpleG\", target_columns = c(\"glia\",\"neurons\") ) #> Training for target 'glia' with 'CimpleG' has finished.: 2.464 sec elapsed #> Training for target 'neurons' with 'CimpleG' has finished.: 0.705 sec elapsed # check signatures cimpleg_result$signatures #> glia neurons #> \"cg14501977\" \"cg24548498\""},{"path":"https://costalab.github.io/CimpleG/reference/compute_ax.html","id":null,"dir":"Reference","previous_headings":"","what":"Feature selection function used in the sigma delta space — compute_ax","title":"Feature selection function used in the sigma delta space — compute_ax","text":"Feature selection function used sigma delta space","code":""},{"path":"https://costalab.github.io/CimpleG/reference/compute_ax.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Feature selection function used in the sigma delta space — compute_ax","text":"","code":"compute_ax(dm, sv, p)"},{"path":"https://costalab.github.io/CimpleG/reference/compute_ax.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Feature selection function used in the sigma delta space — compute_ax","text":"dm, delta (difference mean values) sv, sigma (sum variance values) p, even number, greater 'p' importance given sigma","code":""},{"path":"https://costalab.github.io/CimpleG/reference/compute_diffmeans_sumvar.html","id":null,"dir":"Reference","previous_headings":"","what":"Compute diff mean sum var dataframe — compute_diffmeans_sumvar","title":"Compute diff mean sum var dataframe — compute_diffmeans_sumvar","text":"Compute diff mean sum var dataframe","code":""},{"path":"https://costalab.github.io/CimpleG/reference/compute_diffmeans_sumvar.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compute diff mean sum var dataframe — compute_diffmeans_sumvar","text":"","code":"compute_diffmeans_sumvar(data, target_vector)"},{"path":"https://costalab.github.io/CimpleG/reference/compute_diffmeans_sumvar.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compute diff mean sum var dataframe — compute_diffmeans_sumvar","text":"data Matrix beta values used compute diffmeans sumvar data frame target_vector boolean vector defining samples data part target class","code":""},{"path":"https://costalab.github.io/CimpleG/reference/darken.html","id":null,"dir":"Reference","previous_headings":"","what":"Helper function to darken down a given color. — darken","title":"Helper function to darken down a given color. — darken","text":"Helper function darken given color.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/darken.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Helper function to darken down a given color. — darken","text":"","code":"darken(color, factor = 0.5)"},{"path":"https://costalab.github.io/CimpleG/reference/darken.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Helper function to darken down a given color. — darken","text":"color Color name hex code color factor Multiplicative factor `color` darkened ","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconv_pred_obs_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Scatter plots of observed (true) vs predicted values for deconvolution. — deconv_pred_obs_plot","title":"Scatter plots of observed (true) vs predicted values for deconvolution. — deconv_pred_obs_plot","text":"Produces one plot panel per number methods predictions. plot panel one plot per cell type.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconv_pred_obs_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Scatter plots of observed (true) vs predicted values for deconvolution. — deconv_pred_obs_plot","text":"","code":"deconv_pred_obs_plot( deconv_df, true_values_col, predicted_cols, sample_id_col, group_col, axis_lims = list(x = c(0, 1), y = c(0, 1)) )"},{"path":"https://costalab.github.io/CimpleG/reference/deconv_pred_obs_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Scatter plots of observed (true) vs predicted values for deconvolution. — deconv_pred_obs_plot","text":"deconv_df data.frame meta.data, true values predictions different methods columns. row prediction given sample given group/cell type. true_values_col string name column true values `deconv_df`. predicted_cols vector strings name columns predictions different methods `deconv_df`. sample_id_col string name column sample name ID `deconv_df`. group_col string name column containing cell types groups `deconv_df`. axis_lims list two entries, `x` `y`, defining limits x y axis plot.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconv_ranking_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Boxplot and rankings of deconvolution metrics for deconvolution results. — deconv_ranking_plot","title":"Boxplot and rankings of deconvolution metrics for deconvolution results. — deconv_ranking_plot","text":"Produces data varied deconvolution performance metrics. Produces one boxplot one ranking plot metric.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconv_ranking_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Boxplot and rankings of deconvolution metrics for deconvolution results. — deconv_ranking_plot","text":"","code":"deconv_ranking_plot( deconv_df, true_values_col, predicted_cols, sample_id_col, group_col, metrics = c(\"rmse\", \"r_squared\", \"adj.r.squared\", \"AIC\"), custom_colours = NULL )"},{"path":"https://costalab.github.io/CimpleG/reference/deconv_ranking_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Boxplot and rankings of deconvolution metrics for deconvolution results. — deconv_ranking_plot","text":"deconv_df data.frame meta.data, true values predictions different methods columns. row prediction given sample given group/cell type. true_values_col string name column true values `deconv_df`. predicted_cols vector strings name columns predictions different methods `deconv_df`. sample_id_col string name column sample name ID `deconv_df`. group_col string name column containing cell types groups `deconv_df`. metrics list two entries, `x` `y`, defining limits x y axis plot. custom_colours named vector colours, names values defined `predicted_cols`. `NULL`, default colours used.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_barplot.html","id":null,"dir":"Reference","previous_headings":"","what":"Stacked barplot of deconvolution results — deconvolution_barplot","title":"Stacked barplot of deconvolution results — deconvolution_barplot","text":"Stacked barplot deconvolution results","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_barplot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Stacked barplot of deconvolution results — deconvolution_barplot","text":"","code":"deconvolution_barplot( deconvoluted_data, meta_data, sample_id_column, true_label_column, color_dict = NULL, show_x_label = FALSE, base_size = 14, ... )"},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_barplot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Stacked barplot of deconvolution results — deconvolution_barplot","text":"deconvoluted_data Result running `run_deconvolution` meta_data Data.frame containing metadata deconvoluted samples sample_id_column Name column containing sample id meta_data data.frame true_label_column Name column containing true labels samples meta_data data.frame color_dict Named string featuring colors values labels (true labels) names show_x_label boolean, `TRUE` sample labels X axis shown. Default `FALSE`. base_size integer defining base size text plot. Default `14`. ... Parameters passed ggplot2::theme function.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_barplot.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Stacked barplot of deconvolution results — deconvolution_barplot","text":"list data ggplot2 plot object.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_epidish.html","id":null,"dir":"Reference","previous_headings":"","what":"EpiDISH deconvolution — deconvolution_epidish","title":"EpiDISH deconvolution — deconvolution_epidish","text":"EpiDISH deconvolution","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_epidish.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"EpiDISH deconvolution — deconvolution_epidish","text":"","code":"deconvolution_epidish( ref_mat, new_data, epidish_method = \"CBS\", epidish_nuv = seq(0.1, 1, 0.1), epidish_maxit = 10000 )"},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_epidish.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"EpiDISH deconvolution — deconvolution_epidish","text":"ref_mat Reference matrix. new_data New data matrix. epidish_method One `CBS` (Cibersort), `RPC` (Robust Partial Correlations), `CP` (Constrained Projection). Default `CBS`. See `EpiDISH` documentation information. epidish_nuv vector candidate values used svm. used epidish_method set `CBS`. See `EpiDISH` documentation information. epidish_maxit Integer number max iterations IWLS (Iterative Weighted Least Squares). used epidish_method set `RPC`.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nmf.html","id":null,"dir":"Reference","previous_headings":"","what":"NMF deconvolution — deconvolution_nmf","title":"NMF deconvolution — deconvolution_nmf","text":"NMF deconvolution","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nmf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"NMF deconvolution — deconvolution_nmf","text":"","code":"deconvolution_nmf(weights_mat, values_mat, ...)"},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nmf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"NMF deconvolution — deconvolution_nmf","text":"weights_mat Reference matrix. values_mat New data matrix. ... Extra parameters set NMF options. relevant parameters probably `method` `beta`.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nnls.html","id":null,"dir":"Reference","previous_headings":"","what":"NNLS deconvolution — deconvolution_nnls","title":"NNLS deconvolution — deconvolution_nnls","text":"NNLS deconvolution","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nnls.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"NNLS deconvolution — deconvolution_nnls","text":"","code":"deconvolution_nnls(dt, compute_cols, ref_mat)"},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nnls.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"NNLS deconvolution — deconvolution_nnls","text":"dt data.table new data features/predictions rows samples columns. compute_cols character vector columns deconvolution algorithm ran. ref_mat reference matrix created CimpleG.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/diffmeans_sumvariance_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","text":"Represent CpGs difference means, sum variances space. plot often used select CpGs good classifiers. CpGs often located bottom left bottom right plot.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/diffmeans_sumvariance_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","text":"","code":"diffmeans_sumvariance_plot( data, xcol = \"diff_means\", ycol = \"sum_variance\", feature_id_col = \"id\", is_feature_selected_col = NULL, label_var1 = \"Target\", label_var2 = \"Others\", target_vector = NULL, mean_cutoff = NULL, var_cutoff = NULL, threshold_func = NULL, func_factor = NULL, feats_to_highlight = NULL, cpg_ranking_df = NULL, color_all_points = NULL, pltDensity = TRUE, density_type = c(\"density\", \"histogram\", \"boxplot\", \"violin\", \"densigram\"), plot_dir = NULL, id_tag = NULL, file_tag = NULL, custom_mods = FALSE )"},{"path":"https://costalab.github.io/CimpleG/reference/diffmeans_sumvariance_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","text":"data Data create difference means, sum variances plot. Either data.frame `xcol`,`ycol` `feature_id_col` , `target_vector` `NULL` matrix beta values , given target, difference means target others, sum variances within target others calculated. xcol Column x-axis data ycol Column y-axis data feature_id_col Column feature ID is_feature_selected_col NULL column TRUE/FALSE features highlighted selected label_var1 Label target class label_var2 Label classes target_vector NULL vector target class assignment, see data mean_cutoff numeric draw mean cutoff given position var_cutoff numeric draw variance cutoff given position threshold_func TODO func_factor TODO feats_to_highlight TODO cpg_ranking_df TODO color_all_points TODO pltDensity TODO density_type TODO plot_dir TODO id_tag TODO file_tag TODO custom_mods TODO","code":""},{"path":"https://costalab.github.io/CimpleG/reference/diffmeans_sumvariance_plot.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","text":"","code":"library(\"CimpleG\") # read data data(train_data) data(train_targets) # make basic plot plt <- diffmeans_sumvariance_plot( train_data, target_vector = train_targets$blood_cells == 1 ) print(plt) # make plot with parabola, colored and highlighted features df_dmeansvar <- compute_diffmeans_sumvar( train_data, target_vector = train_targets$blood_cells==1 ) parab_param <- .7 df_dmeansvar$is_selected <- select_features( x = df_dmeansvar$diff_means, y = df_dmeansvar$sum_variance, a = parab_param ) plt <- diffmeans_sumvariance_plot( data=df_dmeansvar, label_var1=\"Leukocytes\", color_all_points=\"red\", is_feature_selected_col=\"is_selected\", feats_to_highlight=c(\"cg10456121\"), threshold_func=function(x,a) (a*x)^2, func_factor=parab_param ) #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Removed 10 rows containing missing values (`geom_line()`). #> Warning: Removed 1 rows containing missing values (`geom_label()`). #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Removed 10 rows containing missing values (`geom_line()`). #> Warning: Removed 1 rows containing missing values (`geom_label()`). #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Removed 10 rows containing missing values (`geom_line()`). #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Removed 10 rows containing missing values (`geom_line()`). #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). print(plt)"},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"Represent CpGs difference means, sum variances space. plot often used select CpGs good classifiers. CpGs often located bottom left bottom right plot.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"","code":"dmsv_plot( dat, target_vector = NULL, x_var = \"diff_means\", y_var = \"sum_variance\", id_var = \"id\", highlight_var = NULL, display_var = NULL, label_var1 = \"Target\", label_var2 = \"Others\", point_color = \"black\", subtitle = NULL )"},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"dat Data create dmsv plot (difference means, sum variances plot). Either data.frame `x_var`,`y_var` `id_var` , `target_vector` `NULL` matrix beta values , given target, difference means target others, sum variances within target others calculated. target_vector NULL boolean vector target class assignment, see data x_var Name column x-axis data (difference means). y_var Name column y-axis data (sum variances). id_var Name column feature/CpG ID. highlight_var (Optional) Name column highlighted features. Values column boolean (TRUE selected, FALSE selected). display_var (Optional) Name column features displayed plot label. Values column boolean (TRUE feature displayed, FALSE feature displayed). label_var1 Label target class. Default \"Target\". label_var2 Label classes. Default \"Others\". point_color Color features/CpGs plot. Default \"black\". features highlighted, non-highlighted features lighter color. subtitle Subtitle displayed plot. Default NULL.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"ggplot2 object dmsv plot.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"","code":"library(\"CimpleG\") # load CimpleG example data data(train_data) data(train_targets) # make basic plot straight from the data plt <- dmsv_plot( dat = train_data, target_vector = train_targets$blood_cells == 1 ) print(plt) # make plot with highlighted features # first create a diffmeans sumvar data frame from the data df_dmeansvar <- compute_diffmeans_sumvar( train_data, target_vector = train_targets$blood_cells==1 ) # adding a column to this data frame \\code{hl_col} with random CpGs # selected (as TRUE) or not (as FALSE) to be highlighted and displayed. df_dmeansvar$hl_col <- sample(c(TRUE,FALSE),nrow(df_dmeansvar),replace=TRUE,prob=c(0.1,0.9)) df_dmeansvar$dp_col <- df_dmeansvar$hl_col plt <- dmsv_plot( dat=df_dmeansvar, highlight_var=\"hl_col\", display_var=\"dp_col\", label_var1=\"Leukocytes\", point_color=\"red\", subtitle=\"method: CimpleG\" ) print(plt) #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). #> Warning: ggrepel: 95 unlabeled data points (too many overlaps). Consider increasing max.overlaps"},{"path":"https://costalab.github.io/CimpleG/reference/eval_test_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluation of produced models on test data — eval_test_data","title":"Evaluation of produced models on test data — eval_test_data","text":"Evaluation produced models test data","code":""},{"path":"https://costalab.github.io/CimpleG/reference/eval_test_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluation of produced models on test data — eval_test_data","text":"","code":"eval_test_data(test_data, final_model, method = \"oner\", verbose = 1)"},{"path":"https://costalab.github.io/CimpleG/reference/eval_test_data.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluation of produced models on test data — eval_test_data","text":"test_data Test data. final_model Model tested. method Method used train model. verbose verbose logs .","code":""},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":null,"dir":"Reference","previous_headings":"","what":"Get CpG annotation from Illumina — get_cpg_annotation","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"Get CpG annotation Illumina","code":""},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"","code":"get_cpg_annotation( cpg_id, is_epic = TRUE, short_annotation = TRUE, silence_warnings = TRUE )"},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"cpg_id character vector CpG IDs Illumina annotate. is_epic boolean, TRUE, annotation fetched EPIC array, otherwise 450k array. Default TRUE. short_annotation boolean, TRUE, small number columns full annotation reference kept. leads easier read output. Default TRUE. silence_warnings boolean, TRUE, warnings produced downloading loading data silenced. Default TRUE.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"table annotated CpGs order provided signatures.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"","code":"library(\"CimpleG\") # read data signatures <- c(\"cg14501977\", \"cg24548498\") # Get signature annotation signature_annotation <- get_cpg_annotation(signatures) #> #> [CimpleG] Getting annotation manifest from Illumina. # check signature annotation signature_annotation #> # A tibble: 2 × 8 #> IlmnID CHR_hg38 Start_hg38 End_hg38 UCSC_RefGene_Name UCSC_RefGene_Group #> #> 1 cg14501977 chr12 123948446 123948448 CCDC92 5'UTR #> 2 cg24548498 chr2 181684680 181684682 NA NA #> # ℹ 2 more variables: UCSC_CpG_Islands_Name , #> # Relation_to_UCSC_CpG_Island "},{"path":"https://costalab.github.io/CimpleG/reference/lighten.html","id":null,"dir":"Reference","previous_headings":"","what":"Helper function to lighten up a given color. — lighten","title":"Helper function to lighten up a given color. — lighten","text":"Helper function lighten given color.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/lighten.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Helper function to lighten up a given color. — lighten","text":"","code":"lighten(color, factor = 0.5)"},{"path":"https://costalab.github.io/CimpleG/reference/lighten.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Helper function to lighten up a given color. — lighten","text":"color Color name hex code color factor Multiplicative factor `color` lightened ","code":""},{"path":"https://costalab.github.io/CimpleG/reference/load_object.html","id":null,"dir":"Reference","previous_headings":"","what":"Load an R object saved with CimpleG or an RDS file. — load_object","title":"Load an R object saved with CimpleG or an RDS file. — load_object","text":"Load R object saved CimpleG RDS file.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/load_object.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load an R object saved with CimpleG or an RDS file. — load_object","text":"","code":"load_object(file_name)"},{"path":"https://costalab.github.io/CimpleG/reference/load_object.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Load an R object saved with CimpleG or an RDS file. — load_object","text":"file_name File name working directory path file loaded. Files saved CimpleG::save_object base::saveRDS files supported.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/make_color_palette.html","id":null,"dir":"Reference","previous_headings":"","what":"Make color palette data frame — make_color_palette","title":"Make color palette data frame — make_color_palette","text":"Make color palette data frame","code":""},{"path":"https://costalab.github.io/CimpleG/reference/make_color_palette.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Make color palette data frame — make_color_palette","text":"","code":"make_color_palette(classes)"},{"path":"https://costalab.github.io/CimpleG/reference/make_color_palette.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Make color palette data frame — make_color_palette","text":"classes Vector classes create color palette","code":""},{"path":"https://costalab.github.io/CimpleG/reference/make_deconv_ref_matrix.html","id":null,"dir":"Reference","previous_headings":"","what":"Build deconvolution reference matrix — make_deconv_ref_matrix","title":"Build deconvolution reference matrix — make_deconv_ref_matrix","text":"Build deconvolution reference matrix","code":""},{"path":"https://costalab.github.io/CimpleG/reference/make_deconv_ref_matrix.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Build deconvolution reference matrix — make_deconv_ref_matrix","text":"","code":"make_deconv_ref_matrix(cpg_obj, ref_data, ref_data_labels, method = NULL)"},{"path":"https://costalab.github.io/CimpleG/reference/make_deconv_ref_matrix.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Build deconvolution reference matrix — make_deconv_ref_matrix","text":"cpg_obj CimpleG object. ref_data matrix reference data used build reference matrix. ref_data_labels character vector true labels samples `reference_data`. method Method used train models CimpleG object. provided (NULL), method taken CimpleG object.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/predict.CimpleG.html","id":null,"dir":"Reference","previous_headings":"","what":"Predict outcome from a CimpleG signatures on new data — predict.CimpleG","title":"Predict outcome from a CimpleG signatures on new data — predict.CimpleG","text":"Predict outcome CimpleG signatures new data","code":""},{"path":"https://costalab.github.io/CimpleG/reference/predict.CimpleG.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Predict outcome from a CimpleG signatures on new data — predict.CimpleG","text":"","code":"# S3 method for CimpleG predict(object, ..., new_data, class_labels = NULL)"},{"path":"https://costalab.github.io/CimpleG/reference/predict.CimpleG.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Predict outcome from a CimpleG signatures on new data — predict.CimpleG","text":"object CimpleG object. ... used moment. new_data Data predicted, samples rows features columns. Last column `new_data` target/class labels coded 0 1. class_labels Class labels new data provided directly .","code":""},{"path":"https://costalab.github.io/CimpleG/reference/run_deconvolution.html","id":null,"dir":"Reference","previous_headings":"","what":"Perform deconvolution on a new set of samples, based on the CimpleG models trained — run_deconvolution","title":"Perform deconvolution on a new set of samples, based on the CimpleG models trained — run_deconvolution","text":"Perform deconvolution new set samples, based CimpleG models trained","code":""},{"path":"https://costalab.github.io/CimpleG/reference/run_deconvolution.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Perform deconvolution on a new set of samples, based on the CimpleG models trained — run_deconvolution","text":"","code":"run_deconvolution( cpg_obj = NULL, new_data = NULL, ref_mat = NULL, deconvolution_method = c(\"NNLS\", \"EpiDISH\", \"NMF\"), ... )"},{"path":"https://costalab.github.io/CimpleG/reference/run_deconvolution.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Perform deconvolution on a new set of samples, based on the CimpleG models trained — run_deconvolution","text":"cpg_obj CimpleG object. creating/training CimpleG parameter `deconvolution_reference` set `TRUE`. new_data Matrix data.frame samples want perform deconvolution . Samples rows probes/CpGs columns. ref_mat CimpleG object reference matrix, can provide instead. See `make_deconv_ref_matrix` deconvolution_method Deconvolution method used. One #TODO ... Extra parameters used deconvolution_method set `NMF`. relevant parameter probably `method` `beta`.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/save_object.html","id":null,"dir":"Reference","previous_headings":"","what":"Save an R object to disk with fast and efficient compression algorithms. — save_object","title":"Save an R object to disk with fast and efficient compression algorithms. — save_object","text":"Save R object disk fast efficient compression algorithms.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/save_object.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Save an R object to disk with fast and efficient compression algorithms. — save_object","text":"","code":"save_object(object, file_name, file_format = \"zstd\")"},{"path":"https://costalab.github.io/CimpleG/reference/save_object.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Save an R object to disk with fast and efficient compression algorithms. — save_object","text":"object Object saved disk. file_name Name file R object saved . file_format One \"zstd\", \"lz4\", \"gzip\", \"bzip2\",\"xz\", \"nocomp\". zstd best option, fast compression loading times, low space usage. Formats \"zstd\" \"lz4\" available package archive installed.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/select_features.html","id":null,"dir":"Reference","previous_headings":"","what":"Feature selection function used in the diffmeans, sumvariance space — select_features","title":"Feature selection function used in the diffmeans, sumvariance space — select_features","text":"Feature selection function used diffmeans, sumvariance space","code":""},{"path":"https://costalab.github.io/CimpleG/reference/select_features.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Feature selection function used in the diffmeans, sumvariance space — select_features","text":"","code":"select_features(x, y, a)"},{"path":"https://costalab.github.io/CimpleG/reference/select_features.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Feature selection function used in the diffmeans, sumvariance space — select_features","text":"x, difference means value y, sum variances value , parabola parameter, scales open/closed parabola , higher value, closed parabola .","code":""},{"path":"https://costalab.github.io/CimpleG/reference/select_features.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Feature selection function used in the diffmeans, sumvariance space — select_features","text":"bool vector","code":""},{"path":"https://costalab.github.io/CimpleG/reference/signature_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"CpG signature plot — signature_plot","title":"CpG signature plot — signature_plot","text":"CpG signature plot","code":""},{"path":"https://costalab.github.io/CimpleG/reference/signature_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"CpG signature plot — signature_plot","text":"","code":"signature_plot( cpg_obj, data, meta_data, sample_id_column, true_label_column, color_dict = NULL, color_others = \"black\", as_panel = TRUE, is_beta = TRUE, base_size = 14, ... )"},{"path":"https://costalab.github.io/CimpleG/reference/signature_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"CpG signature plot — signature_plot","text":"cpg_obj CimpleG object, generated CimpleG function. Alternatively names character vector list signatures. data Matrix data.frame samples signatures plot. Samples rows probes/CpGs columns. meta_data Data.frame containing metadata samples `data`. sample_id_column Name column containing sample id meta_data data.frame true_label_column Name column containing true labels samples meta_data data.frame color_dict Named string featuring colors values labels (true labels) names color_others name hex code color non-target samples colored . as_panel boolean, TRUE (default) single figure panel signatures generated. Otherwise, individual plots returned list. is_beta boolean, TRUE (default) values plotted scale suitable Beta values. Otherwise, values plotted scale suitable M values. base_size integer defining base size text plot. Default `14`. ... Parameters passed ggplot2::theme function.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/signature_plot.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"CpG signature plot — signature_plot","text":"list data ggplot2 plot object.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/test_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Cell line test data — test_data","title":"Cell line test data — test_data","text":"Cell line test data","code":""},{"path":"https://costalab.github.io/CimpleG/reference/test_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cell line test data — test_data","text":"","code":"test_data"},{"path":"https://costalab.github.io/CimpleG/reference/test_data.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Cell line test data — test_data","text":"matrix beta values 1000 CpGs. Features/variables columns 170 samples rows","code":""},{"path":"https://costalab.github.io/CimpleG/reference/test_targets.html","id":null,"dir":"Reference","previous_headings":"","what":"Cell line test data targets — test_targets","title":"Cell line test data targets — test_targets","text":"Cell line test data targets","code":""},{"path":"https://costalab.github.io/CimpleG/reference/test_targets.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cell line test data targets — test_targets","text":"","code":"test_targets"},{"path":"https://costalab.github.io/CimpleG/reference/test_targets.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Cell line test data targets — test_targets","text":"data frame 18 variables 170 samples rows. gsm GSM identifier (GEO accession number) sample cell_type cell type respective sample adipocytes one-hot encoded (1 0) column defining given sample adipocyte astrocytes one-hot encoded (1 0) column defining given sample astrocyte blood_cells one-hot encoded (1 0) column defining given sample blood cell endothelial_cells one-hot encoded (1 0) column defining given sample endothelial cell epidermal_cells one-hot encoded (1 0) column defining given sample epidermal cell epithelial_cells one-hot encoded (1 0) column defining given sample epithelial cell fibroblasts one-hot encoded (1 0) column defining given sample fibroblast glia one-hot encoded (1 0) column defining given sample glia cell hepatocytes one-hot encoded (1 0) column defining given sample hepatocyte ips_cells one-hot encoded (1 0) column defining given sample ipsc msc one-hot encoded (1 0) column defining given sample msc muscle_cells one-hot encoded (1 0) column defining given sample muscle cell neurons one-hot encoded (1 0) column defining given sample neuron muscle_sc one-hot encoded (1 0) column defining given sample muscle stem cell group_data dataset data belong (train test) description cell type respective sample, long form","code":""},{"path":"https://costalab.github.io/CimpleG/reference/train_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Cell line train data — train_data","title":"Cell line train data — train_data","text":"Cell line train data","code":""},{"path":"https://costalab.github.io/CimpleG/reference/train_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cell line train data — train_data","text":"","code":"train_data"},{"path":"https://costalab.github.io/CimpleG/reference/train_data.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Cell line train data — train_data","text":"matrix beta values 1000 CpGs. Features/variables columns 409 samples rows","code":""},{"path":"https://costalab.github.io/CimpleG/reference/train_targets.html","id":null,"dir":"Reference","previous_headings":"","what":"Cell line train data targets — train_targets","title":"Cell line train data targets — train_targets","text":"Cell line train data targets","code":""},{"path":"https://costalab.github.io/CimpleG/reference/train_targets.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cell line train data targets — train_targets","text":"","code":"train_targets"},{"path":"https://costalab.github.io/CimpleG/reference/train_targets.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Cell line train data targets — train_targets","text":"data frame 18 variables 409 samples rows. gsm GSM identifier (GEO accession number) sample cell_type cell type respective sample adipocytes one-hot encoded (1 0) column defining given sample adipocyte astrocytes one-hot encoded (1 0) column defining given sample astrocyte blood_cells one-hot encoded (1 0) column defining given sample blood cell endothelial_cells one-hot encoded (1 0) column defining given sample endothelial cell epidermal_cells one-hot encoded (1 0) column defining given sample epidermal cell epithelial_cells one-hot encoded (1 0) column defining given sample epithelial cell fibroblasts one-hot encoded (1 0) column defining given sample fibroblast glia one-hot encoded (1 0) column defining given sample glia cell hepatocytes one-hot encoded (1 0) column defining given sample hepatocyte ips_cells one-hot encoded (1 0) column defining given sample ipsc msc one-hot encoded (1 0) column defining given sample msc muscle_cells one-hot encoded (1 0) column defining given sample muscle cell neurons one-hot encoded (1 0) column defining given sample neuron muscle_sc one-hot encoded (1 0) column defining given sample muscle stem cell group_data dataset data belong (train test) description cell type respective sample, long form","code":""}]
+[{"path":"https://costalab.github.io/CimpleG/LICENSE.html","id":null,"dir":"","previous_headings":"","what":"GNU General Public License","title":"GNU General Public License","text":"Version 3, 29 June 2007Copyright © 2007 Free Software Foundation, Inc. .","code":" Copyright (C) This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see Copyright (C) This program comes with ABSOLUTELY NO WARRANTY; for details type 'show w'. This is free software, and you are welcome to redistribute it under certain conditions; type 'show c' for details."},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"cimpleg-simple-cpg-signatures","dir":"Articles","previous_headings":"","what":"CimpleG (Simple CpG signatures)","title":"Generate signatures","text":"CimpleG tries find CpGs best classify cell-type given train dataset also enables perform cell-type deconvolution couple easy steps can use beta M values show easy generate signatures","code":""},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"installation","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Installation","title":"Generate signatures","text":"haven’t installed CimpleG, can find instructions . However simple :","code":"if(!require(\"CimpleG\")) devtools::install_github(\"costalab/CimpleG\")"},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"loading-package","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Loading package","title":"Generate signatures","text":"load CimpleG package.","code":"library(\"CimpleG\") #> -------------------------- #> CimpleG version 0.0.5.9001 #> --------------------------"},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"loading-data","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Loading data","title":"Generate signatures","text":"tutorial, use small dataset just 409 samples 1000 CpGs. also use table metadata regarding samples. dataset comes included CimpleG. can read .","code":"# load data data(train_data) data(train_targets)"},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"running-cimpleg","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Running CimpleG","title":"Generate signatures","text":"Running CimpleG can quite simple. just need run CimpleG function parameters. generating signatures find leukocytes hepatocytes.","code":"# run CimpleG cimpleg_result <- CimpleG( train_data, train_targets, target_columns = c(\"blood_cells\",\"hepatocytes\"), train_only = TRUE ) #> Training for target 'blood_cells' with 'CimpleG' has finished.: 2.655 sec elapsed #> Training for target 'hepatocytes' with 'CimpleG' has finished.: 0.482 sec elapsed"},{"path":"https://costalab.github.io/CimpleG/articles/generate-signatures.html","id":"plotting-cimpleg-cpg-signature","dir":"Articles","previous_headings":"CimpleG (Simple CpG signatures)","what":"Plotting CimpleG CpG signature","title":"Generate signatures","text":"can quickly visualize signature able separate data.","code":"signature_plot( cimpleg_result, train_data, train_targets, sample_id_column = \"gsm\", true_label_column = \"cell_type\" ) #> $data #> # A tibble: 818 × 5 #> # Groups: sig_set [2] #> sample_id true_label signatures value sig_set #> #> 1 GSM1415516 adipocytes cg04785083 0.922 blood_cells #> 2 GSM1415516 adipocytes cg02258444 0.938 hepatocytes #> 3 GSM1415518 adipocytes cg04785083 0.932 blood_cells #> 4 GSM1415518 adipocytes cg02258444 0.912 hepatocytes #> 5 GSM1415520 adipocytes cg04785083 0.923 blood_cells #> 6 GSM1415520 adipocytes cg02258444 0.907 hepatocytes #> 7 GSM1415522 adipocytes cg04785083 0.936 blood_cells #> 8 GSM1415522 adipocytes cg02258444 0.908 hepatocytes #> 9 GSM1415526 adipocytes cg04785083 0.938 blood_cells #> 10 GSM1415526 adipocytes cg02258444 0.905 hepatocytes #> # ℹ 808 more rows #> #> $plot"},{"path":"https://costalab.github.io/CimpleG/articles/save_load_objects.html","id":"cimpleg-to-save-and-load-large-files-in-r","dir":"Articles","previous_headings":"","what":"CimpleG to save and load large files in R","title":"Quickly save and load (large) objects","text":"CimpleG comes couple neat functions save load objects using state art data compression algorithms ‘zstd’ ‘lz4’. allows much quickly save load R objects (.rds) files. Furthermore can also use functions load .rds files.","code":""},{"path":"https://costalab.github.io/CimpleG/articles/save_load_objects.html","id":"loading-the-package","dir":"Articles","previous_headings":"CimpleG to save and load large files in R","what":"Loading the package","title":"Quickly save and load (large) objects","text":"load CimpleG package.","code":"library(\"CimpleG\") #> -------------------------- #> CimpleG version 0.0.5.9003 #> --------------------------"},{"path":"https://costalab.github.io/CimpleG/articles/save_load_objects.html","id":"saving-an-object","dir":"Articles","previous_headings":"CimpleG to save and load large files in R","what":"Saving an object","title":"Quickly save and load (large) objects","text":"default use ‘zstd’ algorithm can choose different one. process pretty seamless. Let’s create large matrix save using CimpleG.","code":"my_matrix <- matrix(rnorm(100), nrow = 10, ncol = 10) save_object(my_matrix, \"my_matrix.rds\")"},{"path":"https://costalab.github.io/CimpleG/articles/save_load_objects.html","id":"loading-an-object","dir":"Articles","previous_headings":"CimpleG to save and load large files in R","what":"Loading an object","title":"Quickly save and load (large) objects","text":"load object process just seamless. need specify compression algorithm used. just work! larger file greater advantage using functions.","code":"my_loaded_matrix <- load_object(\"my_matrix.rds\") my_loaded_matrix"},{"path":"https://costalab.github.io/CimpleG/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors and Citation","text":"Tiago F.V. Maié. Author, maintainer.","code":""},{"path":"https://costalab.github.io/CimpleG/authors.html","id":"citation","dir":"","previous_headings":"","what":"Citation","title":"Authors and Citation","text":"F.V. Maié T (2023). CimpleG: CimpleG - Individual/Single CpG sites Classification Deconvolution. R package version 0.0.5.9028, https://costalab.github.io/CimpleG/, https://github.com/CostaLab/CimpleG.","code":"@Manual{, title = {CimpleG: CimpleG - Individual/Single CpG sites For Classification and Deconvolution}, author = {Tiago {F.V. Maié}}, year = {2023}, note = {R package version 0.0.5.9028, https://costalab.github.io/CimpleG/}, url = {https://github.com/CostaLab/CimpleG}, }"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"overview","dir":"","previous_headings":"","what":"Overview","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"CimpleG, R package find (small) CpG signatures.","code":""},{"path":"https://costalab.github.io/CimpleG/index.html","id":"installation","dir":"","previous_headings":"","what":"Installation","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"# Install directly from github: devtools::install_github(\"costalab/CimpleG\") # Alternatively, downloading from our release page and installing it from a local source: # - ie navigating through your system install.packages(file.choose(), repos = NULL, type = \"source\") # - ie given a path to a local source install.packages(\"~/Downloads/CimpleG_0.0.5.XXXX.tar.gz\", repos = NULL, type = \"source\") # or devtools::install_local(\"~/Downloads/CimpleG_0.0.5.XXXX.tar.gz\")"},{"path":"https://costalab.github.io/CimpleG/index.html","id":"getting-started","dir":"","previous_headings":"","what":"Getting started","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"library(\"CimpleG\") data(train_data) data(train_targets) data(test_data) data(test_targets) # check the train_targets table to see # what other columns can be used as targets # colnames(train_targets) # mini example with just 4 target signatures set.seed(42) cimpleg_result <- CimpleG( train_data = train_data, train_targets = train_targets, test_data = test_data, test_targets = test_targets, method = \"CimpleG\", has_annotation = TRUE, target_columns = c( \"neurons\", \"glia\", \"blood_cells\", \"fibroblasts\" ) ) cimpleg_result$results # check generated signatures cimpleg_result$signatures #> neurons glia blood_cells fibroblasts #> \"cg24548498\" \"cg14501977\" \"cg04785083\" \"cg03369247\""},{"path":"https://costalab.github.io/CimpleG/index.html","id":"get-signature-annotation","dir":"","previous_headings":"Getting started","what":"Get signature annotation","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"# Get it directly from the results object cimpleg_result$annotation #> # A tibble: 4 × 8 #> IlmnID CHR_hg38 Start_hg38 End_hg38 UCSC_RefGene_Name UCSC_RefGene_Group #> #> 1 cg24548498 chr2 181684680 181684682 #> 2 cg14501977 chr12 123948446 123948448 CCDC92 5'UTR #> 3 cg04785083 chr1 8971202 8971204 CA6 Body #> 4 cg03369247 chr8 20174518 20174520 SLC18A1;SLC18A1;S… Body;Body;Body;Bo… #> # ℹ 2 more variables: UCSC_CpG_Islands_Name , #> # Relation_to_UCSC_CpG_Island # or idependently through the \"get_cpg_annotation\" function signature_annotation <- get_cpg_annotation(cimpleg_result$signatures) # check signature annotation signature_annotation #> # A tibble: 4 × 8 #> IlmnID CHR_hg38 Start_hg38 End_hg38 UCSC_RefGene_Name UCSC_RefGene_Group #> #> 1 cg24548498 chr2 181684680 181684682 #> 2 cg14501977 chr12 123948446 123948448 CCDC92 5'UTR #> 3 cg04785083 chr1 8971202 8971204 CA6 Body #> 4 cg03369247 chr8 20174518 20174520 SLC18A1;SLC18A1;S… Body;Body;Body;Bo… #> # ℹ 2 more variables: UCSC_CpG_Islands_Name , #> # Relation_to_UCSC_CpG_Island "},{"path":"https://costalab.github.io/CimpleG/index.html","id":"plot-generated-signatures","dir":"","previous_headings":"Getting started","what":"Plot generated signatures","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"# adjust target names to match signature names # check generated signatures plt <- signature_plot( cimpleg_result, train_data, train_targets, sample_id_column = \"gsm\", true_label_column = \"cell_type\" ) print(plt$plot)"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"basic-plot","dir":"","previous_headings":"Difference of means vs Sum of variances (dmsv) plots","what":"basic plot","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"plt <- diffmeans_sumvariance_plot( data = train_data, target_vector = train_targets$neurons == 1 ) print(plt)"},{"path":"https://costalab.github.io/CimpleG/index.html","id":"adding-color-highlighting-selected-features","dir":"","previous_headings":"Difference of means vs Sum of variances (dmsv) plots","what":"adding color, highlighting selected features","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"df_dmeansvar <- compute_diffmeans_sumvar( data = train_data, target_vector = train_targets$neurons == 1 ) parab_param <- .7 df_dmeansvar$is_selected <- select_features( x = df_dmeansvar$diff_means, y = df_dmeansvar$sum_variance, a = parab_param ) plt <- diffmeans_sumvariance_plot( data = df_dmeansvar, label_var1 = \"Neurons\", color_all_points = \"purple\", threshold_func = function(x, a) (a * x) ^ 2, is_feature_selected_col = \"is_selected\", func_factor = parab_param ) print(plt)"},{"path":"https://costalab.github.io/CimpleG/index.html","id":"labeling-specific-features","dir":"","previous_headings":"Difference of means vs Sum of variances (dmsv) plots","what":"labeling specific features","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"plt <- diffmeans_sumvariance_plot( data = df_dmeansvar, feats_to_highlight = cimpleg_result$signatures ) print(plt)"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"mini-example-with-just-4-signatures","dir":"","previous_headings":"Deconvolution plots","what":"mini example with just 4 signatures","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"deconv_result <- run_deconvolution( cpg_obj = cimpleg_result, new_data = test_data ) plt <- deconvolution_barplot( deconvoluted_data = deconv_result, meta_data = test_targets, sample_id = \"gsm\", true_label = \"cell_type\" ) print(plt$plot)"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"first-lets-create-additional-deconvolution-results-so-that-we-can-compare-them","dir":"","previous_headings":"Deconvolution plots > this example is a little more advanced","what":"first lets create additional deconvolution results so that we can compare them","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"example, ’ll create two additional models made CimpleG. One using hypermethylated signatures, using 3 CpGs per signature instead just one.","code":"set.seed(42) cimpleg_hyper <- CimpleG( train_data = train_data, train_targets = train_targets, test_data = test_data, test_targets = test_targets, method = \"CimpleG\", pred_type = \"hyper\", target_columns = c( \"neurons\", \"glia\", \"blood_cells\", \"fibroblasts\" ) ) #> Training for target 'neurons' with 'CimpleG' has finished.: 0.342 sec elapsed #> Training for target 'glia' with 'CimpleG' has finished.: 0.304 sec elapsed #> Training for target 'blood_cells' with 'CimpleG' has finished.: 0.352 sec elapsed #> Training for target 'fibroblasts' with 'CimpleG' has finished.: 0.317 sec elapsed deconv_hyper <- run_deconvolution( cpg_obj = cimpleg_hyper, new_data = test_data ) set.seed(42) cimpleg_3sigs <- CimpleG( train_data = train_data, train_targets = train_targets, test_data = test_data, test_targets = test_targets, method = \"CimpleG\", n_sigs = 3, target_columns = c( \"neurons\", \"glia\", \"blood_cells\", \"fibroblasts\" ) ) #> Training for target 'neurons' with 'CimpleG' has finished.: 0.471 sec elapsed #> Training for target 'glia' with 'CimpleG' has finished.: 0.39 sec elapsed #> Training for target 'blood_cells' with 'CimpleG' has finished.: 0.459 sec elapsed #> Training for target 'fibroblasts' with 'CimpleG' has finished.: 0.403 sec elapsed deconv_3sigs <- run_deconvolution( cpg_obj = cimpleg_3sigs, new_data = test_data )"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"remember-this-is-just-an-example-the-results-themselves-are-meaningless","dir":"","previous_headings":"Deconvolution plots > this example is a little more advanced","what":"remember this is just an example, the results themselves are meaningless!","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"deconv_3sigs$prop_3sigs <- deconv_3sigs$proportion deconv_hyper$prop_hyper <- deconv_hyper$proportion deconv_result$prop_cimpleg <- deconv_result$proportion dummy_deconvolution_data <- deconv_result |> dplyr::mutate(true_vals = proportion + runif(nrow(deconv_result), min=-0.1,max=0.1)) |> dplyr::select(cell_type,sample_id,prop_cimpleg,true_vals) |> dplyr::left_join(deconv_hyper |> dplyr::select(-proportion), by=c(\"sample_id\",\"cell_type\")) |> dplyr::left_join(deconv_3sigs |> dplyr::select(-proportion), by=c(\"sample_id\",\"cell_type\")) |> dplyr::mutate_if(is.numeric, function(x){ifelse(x<0,0,x)}) |> dplyr::mutate_if(is.numeric, function(x){ifelse(x>1,1,x)}) |> tibble::as_tibble()"},{"path":[]},{"path":"https://costalab.github.io/CimpleG/index.html","id":"first-we-can-check-how-the-true-values-compare-against-the-predicted-values","dir":"","previous_headings":"Deconvolution plots > this example is a little more advanced","what":"first we can check how the true values compare against the predicted values","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"scatter_plts <- CimpleG:::deconv_pred_obs_plot( deconv_df = dummy_deconvolution_data, true_values_col = \"true_vals\", predicted_cols = c(\"prop_cimpleg\",\"prop_hyper\",\"prop_3sigs\"), sample_id_col = \"sample_id\", group_col= \"cell_type\" ) scatter_panel <- scatter_plts |> patchwork::wrap_plots(ncol=1) print(scatter_panel)"},{"path":"https://costalab.github.io/CimpleG/index.html","id":"now-more-interestingly-we-can-see-in-detail-and-rank-one-of-the-measures-used-to-evaluate-the-deconvolution-results","dir":"","previous_headings":"Deconvolution plots > this example is a little more advanced","what":"now, more interestingly, we can see in detail and rank one of the measures used to evaluate the deconvolution results","title":"CimpleG - Individual/Single CpG sites For Classification and\n Deconvolution","text":"","code":"rank_plts <- CimpleG:::deconv_ranking_plot( deconv_df = dummy_deconvolution_data, true_values_col = \"true_vals\", predicted_cols = c(\"prop_cimpleg\",\"prop_hyper\",\"prop_3sigs\"), sample_id_col = \"sample_id\", group_col= \"cell_type\", metrics = \"rmse\" ) rank_panel <- list(rank_plts$perf_boxplt[[1]],rank_plts$nemenyi_plt[[1]]) |> patchwork::wrap_plots() print(rank_panel)"},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG-package.html","id":null,"dir":"Reference","previous_headings":"","what":"CimpleG: CimpleG - Individual/Single CpG sites For Classification and Deconvolution — CimpleG-package","title":"CimpleG: CimpleG - Individual/Single CpG sites For Classification and Deconvolution — CimpleG-package","text":"CimpleG simple package find single-CpG based signatures use classification deconvolution.","code":""},{"path":[]},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG-package.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"CimpleG: CimpleG - Individual/Single CpG sites For Classification and Deconvolution — CimpleG-package","text":"Maintainer: Tiago F.V. Maié tiagomaie@hotmail.com","code":""},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":null,"dir":"Reference","previous_headings":"","what":"Find simple CpG (CimpleG) signatures. — CimpleG","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"Train classification model using (CpGs) features given target data.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"","code":"CimpleG( train_data, train_targets = NULL, target_columns = NULL, test_data = NULL, test_targets = NULL, method = c(\"CimpleG\", \"CimpleG_parab\", \"brute_force\", \"logistic_reg\", \"decision_tree\", \"boost_tree\", \"mlp\", \"rand_forest\"), pred_type = c(\"both\", \"hypo\", \"hyper\"), engine = c(\"glmnet\", \"xgboost\", \"nnet\", \"ranger\"), rank_method = c(\"ac_rank\", \"a_rank\", \"c_rank\"), k_folds = 10, grid_n = 10, param_p = 2, n_sigs = 1, quantile_threshold = 0.005, train_only = FALSE, split_data = FALSE, run_parallel = FALSE, deconvolution_reference = TRUE, save_dir = NULL, save_format = c(\"zstd\", \"lz4\", \"gzip\", \"bzip2\", \"xz\", \"nocomp\"), verbose = 1, targets = NULL ) cimpleg( train_data, train_targets = NULL, target_columns = NULL, test_data = NULL, test_targets = NULL, method = c(\"CimpleG\", \"CimpleG_parab\", \"brute_force\", \"logistic_reg\", \"decision_tree\", \"boost_tree\", \"mlp\", \"rand_forest\"), pred_type = c(\"both\", \"hypo\", \"hyper\"), engine = c(\"glmnet\", \"xgboost\", \"nnet\", \"ranger\"), rank_method = c(\"ac_rank\", \"a_rank\", \"c_rank\"), k_folds = 10, grid_n = 10, param_p = 2, n_sigs = 1, quantile_threshold = 0.005, train_only = FALSE, split_data = FALSE, run_parallel = FALSE, deconvolution_reference = TRUE, save_dir = NULL, save_format = c(\"zstd\", \"lz4\", \"gzip\", \"bzip2\", \"xz\", \"nocomp\"), verbose = 1, targets = NULL ) cpg( train_data, train_targets = NULL, target_columns = NULL, test_data = NULL, test_targets = NULL, method = c(\"CimpleG\", \"CimpleG_parab\", \"brute_force\", \"logistic_reg\", \"decision_tree\", \"boost_tree\", \"mlp\", \"rand_forest\"), pred_type = c(\"both\", \"hypo\", \"hyper\"), engine = c(\"glmnet\", \"xgboost\", \"nnet\", \"ranger\"), rank_method = c(\"ac_rank\", \"a_rank\", \"c_rank\"), k_folds = 10, grid_n = 10, param_p = 2, n_sigs = 1, quantile_threshold = 0.005, train_only = FALSE, split_data = FALSE, run_parallel = FALSE, deconvolution_reference = TRUE, save_dir = NULL, save_format = c(\"zstd\", \"lz4\", \"gzip\", \"bzip2\", \"xz\", \"nocomp\"), verbose = 1, targets = NULL )"},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"train_data Training dataset. matrix (s x f) methylation data (Beta values) used train/find predictors. Samples (s) must rows features/CpGs (f) must columns. train_targets data frame training target samples one-hot encoded. data frame least 1 column, many rows order `train_data`. Target columns need one-hot encoded, meaning , column target samples encoded `1` every sample encoded `0`. target_columns string specifying name column `train_targets` used training. Can character vector several columns `train_targets` used training. argument character vector, CimpleG search best predictors target sequentially parallel depending value `run_parallel` test_data Testing dataset. matrix (s x f) methylation data (Beta values) used test performance found predictors. Samples (s) must rows features/CpGs (f) must columns. `test_data` ** `test_targets` NULL, CimpleG generate stratified test dataset based `train_targets` removing 25 samples `train_data` `train_targets`. test_targets data frame testing target samples one-hot encoded. data frame least 1 column, many rows order `test_data`. Target columns need one-hot encoded, meaning , column target samples encoded `1` every sample encoded `0`. `test_data` ** `test_targets` NULL, CimpleG generate stratified test dataset based `train_targets` removing 25 samples `train_data` `train_targets`. method string specifying method type machine learning model/algorithm used training. divided two main groups. * simple models (classifiers use single feature), `CimpleG` (default), `brute_force`, `CimpleG_unscaled` `oner`; * complex models (classifiers use several features), `logistic_reg`, `decision_tree`, `boost_tree`, `mlp` `rand_forest`. pred_type string specifying type predictor/CpG searched training. used simple models. One `` (default), `hypo` `hyper`. `hypo`, hypomethylated predictors considered. `hyper`, hypermethylated predictors considered. engine string specifying machine learning engine behind `method`. used complex models. Currently use. rank_method string specifying ranking strategy rank features training. k_folds integer specifying number folds (K) used training stratified K-fold cross-validation procedure. grid_n integer specifying number hyperparameter combinations train . param_p even number `sigma / (delta^param_p)`. Tunes much weight given delta feature selection. Default 2. n_sigs Number signatures saved classification used deconvolution. Default 1. quantile_threshold number 0 1. Determines many features kept. Default 0.005. train_only boolean, TRUE, CimpleG train (find predictors) test test dataset. split_data boolean, `TRUE`, subset train data provided, creating smaller test set used test models training. parameter experimental. Default `FALSE`. run_parallel boolean, `FALSE`, default, search predictors multiple targets sequentially. `TRUE` search predictors multiple targets time (parallel processing) order save computational time. need set `future::plan()` running function. deconvolution_reference boolean, `TRUE`, create deconvolution reference matrix based training data. can later used perform deconvolution. Default `FALSE`. save_dir defined save resulting model given directory. Default NULL. save_format used save_dir NULL. One \"zstd\", \"lz4\", \"gzip\", \"bzip2\",\"xz\", \"nocomp\". zstd best option, fast compression loading times, low space usage. verbose verbose want CimpleG running. 0, message displayed, 3 every message displayed. Default 1. targets DEPRECATED use `target_columns`.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"CimpleG object results per target class.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/CimpleG.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Find simple CpG (CimpleG) signatures. — CimpleG","text":"","code":"library(\"CimpleG\") # read data data(train_data) data(train_targets) data(test_data) data(test_targets) # run CimpleG cimpleg_result <- CimpleG( train_data = train_data, train_targets = train_targets, test_data = test_data, test_targets = test_targets, method = \"CimpleG\", target_columns = c(\"glia\",\"neurons\") ) #> Training for target 'glia' with 'CimpleG' has finished.: 2.464 sec elapsed #> Training for target 'neurons' with 'CimpleG' has finished.: 0.705 sec elapsed # check signatures cimpleg_result$signatures #> glia neurons #> \"cg14501977\" \"cg24548498\""},{"path":"https://costalab.github.io/CimpleG/reference/compute_ax.html","id":null,"dir":"Reference","previous_headings":"","what":"Feature selection function used in the sigma delta space — compute_ax","title":"Feature selection function used in the sigma delta space — compute_ax","text":"Feature selection function used sigma delta space","code":""},{"path":"https://costalab.github.io/CimpleG/reference/compute_ax.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Feature selection function used in the sigma delta space — compute_ax","text":"","code":"compute_ax(dm, sv, p)"},{"path":"https://costalab.github.io/CimpleG/reference/compute_ax.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Feature selection function used in the sigma delta space — compute_ax","text":"dm, delta (difference mean values) sv, sigma (sum variance values) p, even number, greater 'p' importance given sigma","code":""},{"path":"https://costalab.github.io/CimpleG/reference/compute_diffmeans_sumvar.html","id":null,"dir":"Reference","previous_headings":"","what":"Compute diff mean sum var dataframe — compute_diffmeans_sumvar","title":"Compute diff mean sum var dataframe — compute_diffmeans_sumvar","text":"Compute diff mean sum var dataframe","code":""},{"path":"https://costalab.github.io/CimpleG/reference/compute_diffmeans_sumvar.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compute diff mean sum var dataframe — compute_diffmeans_sumvar","text":"","code":"compute_diffmeans_sumvar(data, target_vector)"},{"path":"https://costalab.github.io/CimpleG/reference/compute_diffmeans_sumvar.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compute diff mean sum var dataframe — compute_diffmeans_sumvar","text":"data Matrix beta values used compute diffmeans sumvar data frame target_vector boolean vector defining samples data part target class","code":""},{"path":"https://costalab.github.io/CimpleG/reference/darken.html","id":null,"dir":"Reference","previous_headings":"","what":"Helper function to darken down a given color. — darken","title":"Helper function to darken down a given color. — darken","text":"Helper function darken given color.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/darken.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Helper function to darken down a given color. — darken","text":"","code":"darken(color, factor = 0.5)"},{"path":"https://costalab.github.io/CimpleG/reference/darken.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Helper function to darken down a given color. — darken","text":"color Color name hex code color factor Multiplicative factor `color` darkened ","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconv_pred_obs_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Scatter plots of observed (true) vs predicted values for deconvolution. — deconv_pred_obs_plot","title":"Scatter plots of observed (true) vs predicted values for deconvolution. — deconv_pred_obs_plot","text":"Produces one plot panel per number methods predictions. plot panel one plot per cell type.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconv_pred_obs_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Scatter plots of observed (true) vs predicted values for deconvolution. — deconv_pred_obs_plot","text":"","code":"deconv_pred_obs_plot( deconv_df, true_values_col, predicted_cols, sample_id_col, group_col, axis_lims = list(x = c(0, 1), y = c(0, 1)) )"},{"path":"https://costalab.github.io/CimpleG/reference/deconv_pred_obs_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Scatter plots of observed (true) vs predicted values for deconvolution. — deconv_pred_obs_plot","text":"deconv_df data.frame meta.data, true values predictions different methods columns. row prediction given sample given group/cell type. true_values_col string name column true values `deconv_df`. predicted_cols vector strings name columns predictions different methods `deconv_df`. sample_id_col string name column sample name ID `deconv_df`. group_col string name column containing cell types groups `deconv_df`. axis_lims list two entries, `x` `y`, defining limits x y axis plot.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconv_ranking_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Boxplot and rankings of deconvolution metrics for deconvolution results. — deconv_ranking_plot","title":"Boxplot and rankings of deconvolution metrics for deconvolution results. — deconv_ranking_plot","text":"Produces data varied deconvolution performance metrics. Produces one boxplot one ranking plot metric.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconv_ranking_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Boxplot and rankings of deconvolution metrics for deconvolution results. — deconv_ranking_plot","text":"","code":"deconv_ranking_plot( deconv_df, true_values_col, predicted_cols, sample_id_col, group_col, metrics = c(\"rmse\", \"r_squared\", \"adj.r.squared\", \"AIC\"), custom_colours = NULL )"},{"path":"https://costalab.github.io/CimpleG/reference/deconv_ranking_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Boxplot and rankings of deconvolution metrics for deconvolution results. — deconv_ranking_plot","text":"deconv_df data.frame meta.data, true values predictions different methods columns. row prediction given sample given group/cell type. true_values_col string name column true values `deconv_df`. predicted_cols vector strings name columns predictions different methods `deconv_df`. sample_id_col string name column sample name ID `deconv_df`. group_col string name column containing cell types groups `deconv_df`. metrics list two entries, `x` `y`, defining limits x y axis plot. custom_colours named vector colours, names values defined `predicted_cols`. `NULL`, default colours used.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_barplot.html","id":null,"dir":"Reference","previous_headings":"","what":"Stacked barplot of deconvolution results — deconvolution_barplot","title":"Stacked barplot of deconvolution results — deconvolution_barplot","text":"Stacked barplot deconvolution results","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_barplot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Stacked barplot of deconvolution results — deconvolution_barplot","text":"","code":"deconvolution_barplot( deconvoluted_data, meta_data, sample_id_column, true_label_column, color_dict = NULL, show_x_label = FALSE, base_size = 14, ... )"},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_barplot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Stacked barplot of deconvolution results — deconvolution_barplot","text":"deconvoluted_data Result running `run_deconvolution` meta_data Data.frame containing metadata deconvoluted samples sample_id_column Name column containing sample id meta_data data.frame true_label_column Name column containing true labels samples meta_data data.frame color_dict Named string featuring colors values labels (true labels) names show_x_label boolean, `TRUE` sample labels X axis shown. Default `FALSE`. base_size integer defining base size text plot. Default `14`. ... Parameters passed ggplot2::theme function.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_barplot.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Stacked barplot of deconvolution results — deconvolution_barplot","text":"list data ggplot2 plot object.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_epidish.html","id":null,"dir":"Reference","previous_headings":"","what":"EpiDISH deconvolution — deconvolution_epidish","title":"EpiDISH deconvolution — deconvolution_epidish","text":"EpiDISH deconvolution","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_epidish.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"EpiDISH deconvolution — deconvolution_epidish","text":"","code":"deconvolution_epidish( ref_mat, new_data, epidish_method = \"CBS\", epidish_nuv = seq(0.1, 1, 0.1), epidish_maxit = 10000 )"},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_epidish.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"EpiDISH deconvolution — deconvolution_epidish","text":"ref_mat Reference matrix. new_data New data matrix. epidish_method One `CBS` (Cibersort), `RPC` (Robust Partial Correlations), `CP` (Constrained Projection). Default `CBS`. See `EpiDISH` documentation information. epidish_nuv vector candidate values used svm. used epidish_method set `CBS`. See `EpiDISH` documentation information. epidish_maxit Integer number max iterations IWLS (Iterative Weighted Least Squares). used epidish_method set `RPC`.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nmf.html","id":null,"dir":"Reference","previous_headings":"","what":"NMF deconvolution — deconvolution_nmf","title":"NMF deconvolution — deconvolution_nmf","text":"NMF deconvolution","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nmf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"NMF deconvolution — deconvolution_nmf","text":"","code":"deconvolution_nmf(weights_mat, values_mat, ...)"},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nmf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"NMF deconvolution — deconvolution_nmf","text":"weights_mat Reference matrix. values_mat New data matrix. ... Extra parameters set NMF options. relevant parameters probably `method` `beta`.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nnls.html","id":null,"dir":"Reference","previous_headings":"","what":"NNLS deconvolution — deconvolution_nnls","title":"NNLS deconvolution — deconvolution_nnls","text":"NNLS deconvolution","code":""},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nnls.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"NNLS deconvolution — deconvolution_nnls","text":"","code":"deconvolution_nnls(dt, compute_cols, ref_mat)"},{"path":"https://costalab.github.io/CimpleG/reference/deconvolution_nnls.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"NNLS deconvolution — deconvolution_nnls","text":"dt data.table new data features/predictions rows samples columns. compute_cols character vector columns deconvolution algorithm ran. ref_mat reference matrix created CimpleG.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/diffmeans_sumvariance_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","text":"Represent CpGs difference means, sum variances space. plot often used select CpGs good classifiers. CpGs often located bottom left bottom right plot.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/diffmeans_sumvariance_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","text":"","code":"diffmeans_sumvariance_plot( data, xcol = \"diff_means\", ycol = \"sum_variance\", feature_id_col = \"id\", is_feature_selected_col = NULL, label_var1 = \"Target\", label_var2 = \"Others\", target_vector = NULL, mean_cutoff = NULL, var_cutoff = NULL, threshold_func = NULL, func_factor = NULL, feats_to_highlight = NULL, cpg_ranking_df = NULL, color_all_points = NULL, pltDensity = TRUE, density_type = c(\"density\", \"histogram\", \"boxplot\", \"violin\", \"densigram\"), plot_dir = NULL, id_tag = NULL, file_tag = NULL, custom_mods = FALSE )"},{"path":"https://costalab.github.io/CimpleG/reference/diffmeans_sumvariance_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","text":"data Data create difference means, sum variances plot. Either data.frame `xcol`,`ycol` `feature_id_col` , `target_vector` `NULL` matrix beta values , given target, difference means target others, sum variances within target others calculated. xcol Column x-axis data ycol Column y-axis data feature_id_col Column feature ID is_feature_selected_col NULL column TRUE/FALSE features highlighted selected label_var1 Label target class label_var2 Label classes target_vector NULL vector target class assignment, see data mean_cutoff numeric draw mean cutoff given position var_cutoff numeric draw variance cutoff given position threshold_func TODO func_factor TODO feats_to_highlight TODO cpg_ranking_df TODO color_all_points TODO pltDensity TODO density_type TODO plot_dir TODO id_tag TODO file_tag TODO custom_mods TODO","code":""},{"path":"https://costalab.github.io/CimpleG/reference/diffmeans_sumvariance_plot.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — diffmeans_sumvariance_plot","text":"","code":"library(\"CimpleG\") # read data data(train_data) data(train_targets) # make basic plot plt <- diffmeans_sumvariance_plot( train_data, target_vector = train_targets$blood_cells == 1 ) print(plt) # make plot with parabola, colored and highlighted features df_dmeansvar <- compute_diffmeans_sumvar( train_data, target_vector = train_targets$blood_cells==1 ) parab_param <- .7 df_dmeansvar$is_selected <- select_features( x = df_dmeansvar$diff_means, y = df_dmeansvar$sum_variance, a = parab_param ) plt <- diffmeans_sumvariance_plot( data=df_dmeansvar, label_var1=\"Leukocytes\", color_all_points=\"red\", is_feature_selected_col=\"is_selected\", feats_to_highlight=c(\"cg10456121\"), threshold_func=function(x,a) (a*x)^2, func_factor=parab_param ) #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Removed 10 rows containing missing values (`geom_line()`). #> Warning: Removed 1 rows containing missing values (`geom_label()`). #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Removed 10 rows containing missing values (`geom_line()`). #> Warning: Removed 1 rows containing missing values (`geom_label()`). #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Removed 10 rows containing missing values (`geom_line()`). #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). #> Warning: Use of `sp_df$x` is discouraged. #> ℹ Use `x` instead. #> Warning: Removed 10 rows containing missing values (`geom_line()`). #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). print(plt)"},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"Represent CpGs difference means, sum variances space. plot often used select CpGs good classifiers. CpGs often located bottom left bottom right plot.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"","code":"dmsv_plot( dat, target_vector = NULL, x_var = \"diff_means\", y_var = \"sum_variance\", id_var = \"id\", highlight_var = NULL, display_var = NULL, label_var1 = \"Target\", label_var2 = \"Others\", point_color = \"black\", subtitle = NULL )"},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"dat Data create dmsv plot (difference means, sum variances plot). Either data.frame `x_var`,`y_var` `id_var` , `target_vector` `NULL` matrix beta values , given target, difference means target others, sum variances within target others calculated. target_vector NULL boolean vector target class assignment, see data x_var Name column x-axis data (difference means). y_var Name column y-axis data (sum variances). id_var Name column feature/CpG ID. highlight_var (Optional) Name column highlighted features. Values column boolean (TRUE selected, FALSE selected). display_var (Optional) Name column features displayed plot label. Values column boolean (TRUE feature displayed, FALSE feature displayed). label_var1 Label target class. Default \"Target\". label_var2 Label classes. Default \"Others\". point_color Color features/CpGs plot. Default \"black\". features highlighted, non-highlighted features lighter color. subtitle Subtitle displayed plot. Default NULL.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"ggplot2 object dmsv plot.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/dmsv_plot.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Represent CpGs in the difference in means, sum of variances space.\nThis plot is often used to select CpGs that would be good classifiers.\nThese CpGs are often located on the bottom left and\nbottom right of this plot. — dmsv_plot","text":"","code":"library(\"CimpleG\") # load CimpleG example data data(train_data) data(train_targets) # make basic plot straight from the data plt <- dmsv_plot( dat = train_data, target_vector = train_targets$blood_cells == 1 ) print(plt) # make plot with highlighted features # first create a diffmeans sumvar data frame from the data df_dmeansvar <- compute_diffmeans_sumvar( train_data, target_vector = train_targets$blood_cells==1 ) # adding a column to this data frame \\code{hl_col} with random CpGs # selected (as TRUE) or not (as FALSE) to be highlighted and displayed. df_dmeansvar$hl_col <- sample(c(TRUE,FALSE),nrow(df_dmeansvar),replace=TRUE,prob=c(0.1,0.9)) df_dmeansvar$dp_col <- df_dmeansvar$hl_col plt <- dmsv_plot( dat=df_dmeansvar, highlight_var=\"hl_col\", display_var=\"dp_col\", label_var1=\"Leukocytes\", point_color=\"red\", subtitle=\"method: CimpleG\" ) print(plt) #> Warning: Removed 1 rows containing missing values (`geom_label_repel()`). #> Warning: ggrepel: 95 unlabeled data points (too many overlaps). Consider increasing max.overlaps"},{"path":"https://costalab.github.io/CimpleG/reference/eval_test_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluation of produced models on test data — eval_test_data","title":"Evaluation of produced models on test data — eval_test_data","text":"Evaluation produced models test data","code":""},{"path":"https://costalab.github.io/CimpleG/reference/eval_test_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluation of produced models on test data — eval_test_data","text":"","code":"eval_test_data(test_data, final_model, method = \"oner\", verbose = 1)"},{"path":"https://costalab.github.io/CimpleG/reference/eval_test_data.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluation of produced models on test data — eval_test_data","text":"test_data Test data. final_model Model tested. method Method used train model. verbose verbose logs .","code":""},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":null,"dir":"Reference","previous_headings":"","what":"Get CpG annotation from Illumina — get_cpg_annotation","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"Get CpG annotation Illumina","code":""},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"","code":"get_cpg_annotation( cpg_id, is_epic = TRUE, short_annotation = TRUE, silence_warnings = TRUE )"},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"cpg_id character vector CpG IDs Illumina annotate. is_epic boolean, TRUE, annotation fetched EPIC array, otherwise 450k array. Default TRUE. short_annotation boolean, TRUE, small number columns full annotation reference kept. leads easier read output. Default TRUE. silence_warnings boolean, TRUE, warnings produced downloading loading data silenced. Default TRUE.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"table annotated CpGs order provided signatures.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/get_cpg_annotation.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Get CpG annotation from Illumina — get_cpg_annotation","text":"","code":"library(\"CimpleG\") # read data signatures <- c(\"cg14501977\", \"cg24548498\") # Get signature annotation signature_annotation <- get_cpg_annotation(signatures) #> #> [CimpleG] Getting annotation manifest from Illumina. # check signature annotation signature_annotation #> # A tibble: 2 × 8 #> IlmnID CHR_hg38 Start_hg38 End_hg38 UCSC_RefGene_Name UCSC_RefGene_Group #> #> 1 cg14501977 chr12 123948446 123948448 CCDC92 5'UTR #> 2 cg24548498 chr2 181684680 181684682 NA NA #> # ℹ 2 more variables: UCSC_CpG_Islands_Name , #> # Relation_to_UCSC_CpG_Island "},{"path":"https://costalab.github.io/CimpleG/reference/lighten.html","id":null,"dir":"Reference","previous_headings":"","what":"Helper function to lighten up a given color. — lighten","title":"Helper function to lighten up a given color. — lighten","text":"Helper function lighten given color.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/lighten.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Helper function to lighten up a given color. — lighten","text":"","code":"lighten(color, factor = 0.5)"},{"path":"https://costalab.github.io/CimpleG/reference/lighten.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Helper function to lighten up a given color. — lighten","text":"color Color name hex code color factor Multiplicative factor `color` lightened ","code":""},{"path":"https://costalab.github.io/CimpleG/reference/load_object.html","id":null,"dir":"Reference","previous_headings":"","what":"Load an R object saved with CimpleG or an RDS file. — load_object","title":"Load an R object saved with CimpleG or an RDS file. — load_object","text":"Load R object saved CimpleG RDS file.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/load_object.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load an R object saved with CimpleG or an RDS file. — load_object","text":"","code":"load_object(file_name)"},{"path":"https://costalab.github.io/CimpleG/reference/load_object.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Load an R object saved with CimpleG or an RDS file. — load_object","text":"file_name File name working directory path file loaded. Files saved CimpleG::save_object base::saveRDS files supported.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/make_color_palette.html","id":null,"dir":"Reference","previous_headings":"","what":"Make color palette data frame — make_color_palette","title":"Make color palette data frame — make_color_palette","text":"Make color palette data frame","code":""},{"path":"https://costalab.github.io/CimpleG/reference/make_color_palette.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Make color palette data frame — make_color_palette","text":"","code":"make_color_palette(classes)"},{"path":"https://costalab.github.io/CimpleG/reference/make_color_palette.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Make color palette data frame — make_color_palette","text":"classes Vector classes create color palette","code":""},{"path":"https://costalab.github.io/CimpleG/reference/make_deconv_ref_matrix.html","id":null,"dir":"Reference","previous_headings":"","what":"Build deconvolution reference matrix — make_deconv_ref_matrix","title":"Build deconvolution reference matrix — make_deconv_ref_matrix","text":"Build deconvolution reference matrix","code":""},{"path":"https://costalab.github.io/CimpleG/reference/make_deconv_ref_matrix.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Build deconvolution reference matrix — make_deconv_ref_matrix","text":"","code":"make_deconv_ref_matrix(cpg_obj, ref_data, ref_data_labels, method = NULL)"},{"path":"https://costalab.github.io/CimpleG/reference/make_deconv_ref_matrix.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Build deconvolution reference matrix — make_deconv_ref_matrix","text":"cpg_obj CimpleG object. ref_data matrix reference data used build reference matrix. ref_data_labels character vector true labels samples `reference_data`. method Method used train models CimpleG object. provided (NULL), method taken CimpleG object.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/predict.CimpleG.html","id":null,"dir":"Reference","previous_headings":"","what":"Predict outcome from a CimpleG signatures on new data — predict.CimpleG","title":"Predict outcome from a CimpleG signatures on new data — predict.CimpleG","text":"Predict outcome CimpleG signatures new data","code":""},{"path":"https://costalab.github.io/CimpleG/reference/predict.CimpleG.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Predict outcome from a CimpleG signatures on new data — predict.CimpleG","text":"","code":"# S3 method for CimpleG predict(object, ..., new_data, class_labels = NULL)"},{"path":"https://costalab.github.io/CimpleG/reference/predict.CimpleG.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Predict outcome from a CimpleG signatures on new data — predict.CimpleG","text":"object CimpleG object. ... used moment. new_data Data predicted, samples rows features columns. Last column `new_data` target/class labels coded 0 1. class_labels Class labels new data provided directly .","code":""},{"path":"https://costalab.github.io/CimpleG/reference/run_deconvolution.html","id":null,"dir":"Reference","previous_headings":"","what":"Perform deconvolution on a new set of samples, based on the CimpleG models trained — run_deconvolution","title":"Perform deconvolution on a new set of samples, based on the CimpleG models trained — run_deconvolution","text":"Perform deconvolution new set samples, based CimpleG models trained","code":""},{"path":"https://costalab.github.io/CimpleG/reference/run_deconvolution.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Perform deconvolution on a new set of samples, based on the CimpleG models trained — run_deconvolution","text":"","code":"run_deconvolution( cpg_obj = NULL, new_data = NULL, ref_mat = NULL, deconvolution_method = c(\"NNLS\", \"EpiDISH\", \"NMF\"), ... )"},{"path":"https://costalab.github.io/CimpleG/reference/run_deconvolution.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Perform deconvolution on a new set of samples, based on the CimpleG models trained — run_deconvolution","text":"cpg_obj CimpleG object. creating/training CimpleG parameter `deconvolution_reference` set `TRUE`. new_data Matrix data.frame samples want perform deconvolution . Samples rows probes/CpGs columns. ref_mat CimpleG object reference matrix, can provide instead. See `make_deconv_ref_matrix` deconvolution_method Deconvolution method used. One #TODO ... Extra parameters used deconvolution_method set `NMF`. relevant parameter probably `method` `beta`.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/save_object.html","id":null,"dir":"Reference","previous_headings":"","what":"Save an R object to disk with fast and efficient compression algorithms. — save_object","title":"Save an R object to disk with fast and efficient compression algorithms. — save_object","text":"Save R object disk fast efficient compression algorithms.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/save_object.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Save an R object to disk with fast and efficient compression algorithms. — save_object","text":"","code":"save_object(object, file_name, file_format = \"zstd\")"},{"path":"https://costalab.github.io/CimpleG/reference/save_object.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Save an R object to disk with fast and efficient compression algorithms. — save_object","text":"object Object saved disk. file_name Name file R object saved . file_format One \"zstd\", \"lz4\", \"gzip\", \"bzip2\",\"xz\", \"nocomp\". zstd best option, fast compression loading times, low space usage. Formats \"zstd\" \"lz4\" available package archive installed.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/select_features.html","id":null,"dir":"Reference","previous_headings":"","what":"Feature selection function used in the diffmeans, sumvariance space — select_features","title":"Feature selection function used in the diffmeans, sumvariance space — select_features","text":"Feature selection function used diffmeans, sumvariance space","code":""},{"path":"https://costalab.github.io/CimpleG/reference/select_features.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Feature selection function used in the diffmeans, sumvariance space — select_features","text":"","code":"select_features(x, y, a)"},{"path":"https://costalab.github.io/CimpleG/reference/select_features.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Feature selection function used in the diffmeans, sumvariance space — select_features","text":"x, difference means value y, sum variances value , parabola parameter, scales open/closed parabola , higher value, closed parabola .","code":""},{"path":"https://costalab.github.io/CimpleG/reference/select_features.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Feature selection function used in the diffmeans, sumvariance space — select_features","text":"bool vector","code":""},{"path":"https://costalab.github.io/CimpleG/reference/signature_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"CpG signature plot — signature_plot","title":"CpG signature plot — signature_plot","text":"CpG signature plot","code":""},{"path":"https://costalab.github.io/CimpleG/reference/signature_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"CpG signature plot — signature_plot","text":"","code":"signature_plot( cpg_obj, data, meta_data, sample_id_column, true_label_column, color_dict = NULL, color_others = \"black\", as_panel = TRUE, is_beta = TRUE, base_size = 14, ... )"},{"path":"https://costalab.github.io/CimpleG/reference/signature_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"CpG signature plot — signature_plot","text":"cpg_obj CimpleG object, generated CimpleG function. Alternatively names character vector list signatures. data Matrix data.frame samples signatures plot. Samples rows probes/CpGs columns. meta_data Data.frame containing metadata samples `data`. sample_id_column Name column containing sample id meta_data data.frame true_label_column Name column containing true labels samples meta_data data.frame color_dict Named string featuring colors values labels (true labels) names color_others name hex code color non-target samples colored . as_panel boolean, TRUE (default) single figure panel signatures generated. Otherwise, individual plots returned list. is_beta boolean, TRUE (default) values plotted scale suitable Beta values. Otherwise, values plotted scale suitable M values. base_size integer defining base size text plot. Default `14`. ... Parameters passed ggplot2::theme function.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/signature_plot.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"CpG signature plot — signature_plot","text":"list data ggplot2 plot object.","code":""},{"path":"https://costalab.github.io/CimpleG/reference/test_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Cell line test data — test_data","title":"Cell line test data — test_data","text":"Cell line test data","code":""},{"path":"https://costalab.github.io/CimpleG/reference/test_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cell line test data — test_data","text":"","code":"test_data"},{"path":"https://costalab.github.io/CimpleG/reference/test_data.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Cell line test data — test_data","text":"matrix beta values 1000 CpGs. Features/variables columns 170 samples rows","code":""},{"path":"https://costalab.github.io/CimpleG/reference/test_targets.html","id":null,"dir":"Reference","previous_headings":"","what":"Cell line test data targets — test_targets","title":"Cell line test data targets — test_targets","text":"Cell line test data targets","code":""},{"path":"https://costalab.github.io/CimpleG/reference/test_targets.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cell line test data targets — test_targets","text":"","code":"test_targets"},{"path":"https://costalab.github.io/CimpleG/reference/test_targets.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Cell line test data targets — test_targets","text":"data frame 18 variables 170 samples rows. gsm GSM identifier (GEO accession number) sample cell_type cell type respective sample adipocytes one-hot encoded (1 0) column defining given sample adipocyte astrocytes one-hot encoded (1 0) column defining given sample astrocyte blood_cells one-hot encoded (1 0) column defining given sample blood cell endothelial_cells one-hot encoded (1 0) column defining given sample endothelial cell epidermal_cells one-hot encoded (1 0) column defining given sample epidermal cell epithelial_cells one-hot encoded (1 0) column defining given sample epithelial cell fibroblasts one-hot encoded (1 0) column defining given sample fibroblast glia one-hot encoded (1 0) column defining given sample glia cell hepatocytes one-hot encoded (1 0) column defining given sample hepatocyte ips_cells one-hot encoded (1 0) column defining given sample ipsc msc one-hot encoded (1 0) column defining given sample msc muscle_cells one-hot encoded (1 0) column defining given sample muscle cell neurons one-hot encoded (1 0) column defining given sample neuron muscle_sc one-hot encoded (1 0) column defining given sample muscle stem cell group_data dataset data belong (train test) description cell type respective sample, long form","code":""},{"path":"https://costalab.github.io/CimpleG/reference/train_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Cell line train data — train_data","title":"Cell line train data — train_data","text":"Cell line train data","code":""},{"path":"https://costalab.github.io/CimpleG/reference/train_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cell line train data — train_data","text":"","code":"train_data"},{"path":"https://costalab.github.io/CimpleG/reference/train_data.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Cell line train data — train_data","text":"matrix beta values 1000 CpGs. Features/variables columns 409 samples rows","code":""},{"path":"https://costalab.github.io/CimpleG/reference/train_targets.html","id":null,"dir":"Reference","previous_headings":"","what":"Cell line train data targets — train_targets","title":"Cell line train data targets — train_targets","text":"Cell line train data targets","code":""},{"path":"https://costalab.github.io/CimpleG/reference/train_targets.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cell line train data targets — train_targets","text":"","code":"train_targets"},{"path":"https://costalab.github.io/CimpleG/reference/train_targets.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Cell line train data targets — train_targets","text":"data frame 18 variables 409 samples rows. gsm GSM identifier (GEO accession number) sample cell_type cell type respective sample adipocytes one-hot encoded (1 0) column defining given sample adipocyte astrocytes one-hot encoded (1 0) column defining given sample astrocyte blood_cells one-hot encoded (1 0) column defining given sample blood cell endothelial_cells one-hot encoded (1 0) column defining given sample endothelial cell epidermal_cells one-hot encoded (1 0) column defining given sample epidermal cell epithelial_cells one-hot encoded (1 0) column defining given sample epithelial cell fibroblasts one-hot encoded (1 0) column defining given sample fibroblast glia one-hot encoded (1 0) column defining given sample glia cell hepatocytes one-hot encoded (1 0) column defining given sample hepatocyte ips_cells one-hot encoded (1 0) column defining given sample ipsc msc one-hot encoded (1 0) column defining given sample msc muscle_cells one-hot encoded (1 0) column defining given sample muscle cell neurons one-hot encoded (1 0) column defining given sample neuron muscle_sc one-hot encoded (1 0) column defining given sample muscle stem cell group_data dataset data belong (train test) description cell type respective sample, long form","code":""}]