DataTrack - undesired behavior when using transformation and windowing

[jayoung]

hi there,

I'm loving Gviz - thanks for this package!

This isn't exactly a bug report - more a request to consider handling transformation/windowing a bit differently.

In DataTrack I can see that the data gets transformed BEFORE it gets windowed. If we use a transformation of log10, just to plot on a log scale, this leads to undesired outcomes (because log10(a) + log10(b) != log10(a+b)). I think what most people would want is simply to have the same plot on a log scale, rather than a different-looking plot.

I figured out a way to get what I want (using aggregation=function(x) {log10(sum(x))}) but maybe it would also make sense to change the underlying DataTrack code to transform AFTER aggregating over windows. Kind of depends on what typical use case for transformation is, but I would think this will be a very common one.

Here's some example code to demonstrate is below.

What do you think?

thanks very much,

Janet Young
Fred Hutch Cancer Center

library(Gviz)

## set up data
# middle position has two overlapping regions, but all positions sum to the same count. 
# Looks fine on a linear scale but not on a log scale
test_sites <- GRanges(seqnames="chr1",
                      ranges=IRanges(start=c(10,20,20,30), width=1),
                      counts=c(100,50,50,100)) 
test_region <- GRanges(seqnames="chr1",
                       ranges=IRanges(start=1, end=40)) 

## axis track
myTrack1 <- GenomeAxisTrack(test_region)

## raw counts track
myTrack2a <- DataTrack(
    test_sites, 
    name="counts",
    region=test_region,
    type = c("h"),
    window=width(test_region), 
    aggregation="sum",
    ylim=c(0,150)
)

## use transformation option to log10 scale - the plot shows that the transformation is occurring BEFORE the window aggregation.  This isn't just like plotting the data on a log scale
myTrack2b <- DataTrack(
    test_sites, 
    name="log10 counts v1 ",
    region=test_region,
    type = c("h"),
    window=width(test_region), 
    aggregation="sum",
    transformation=function(x) {log10(x)},
    ylim=c(0,3.5)
)

## SOLUTION - do the log10 scale transformation during window aggregation instead, so we can control the order of aggregation and transformation:
myTrack2c <- DataTrack(
    test_sites, 
    name="log10 counts v2",
    region=test_region,
    type = c("h"),
![gviz_transformation_question](https://github.com/ivanek/Gviz/assets/2569397/0c6c3f78-fb98-48b9-bc3e-500c02fe8899)
[](url)
    window=width(test_region), 
    aggregation=function(x) {log10(sum(x))},
    ylim=c(0,3.5)
)

myTrackList <- list(myTrack2c,myTrack2b,myTrack2a,myTrack1)

png(filename="bug_testing/gviz_transformation_question.png")
plotTracks(myTrackList)
dev.off()

[jayoung]

[jayoung]

sessionInfo()

R version 4.3.2 (2023-10-31)
Platform: x86_64-apple-darwin20 (64-bit)
Running under: macOS Ventura 13.6.6

Matrix products: default
BLAS:   /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib 
LAPACK: /Library/Frameworks/R.framework/Versions/4.3-x86_64/Resources/lib/libRlapack.dylib;  LAPACK version 3.11.0

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

time zone: America/Los_Angeles
tzcode source: internal

attached base packages:
[1] grid      stats4    stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] Gviz_1.46.1          GenomicRanges_1.54.1 GenomeInfoDb_1.38.8  IRanges_2.36.0      
[5] S4Vectors_0.40.2     BiocGenerics_0.48.1 

loaded via a namespace (and not attached):
  [1] DBI_1.2.2                   bitops_1.0-7                deldir_2.0-4               
  [4] gridExtra_2.3               biomaRt_2.58.2              rlang_1.1.3                
  [7] magrittr_2.0.3              biovizBase_1.50.0           matrixStats_1.3.0          
 [10] compiler_4.3.2              RSQLite_2.3.6               GenomicFeatures_1.54.4     
 [13] png_0.1-8                   vctrs_0.6.5                 ProtGenerics_1.34.0        
 [16] stringr_1.5.1               pkgconfig_2.0.3             crayon_1.5.2               
 [19] fastmap_1.1.1               backports_1.4.1             dbplyr_2.5.0               
 [22] XVector_0.42.0              utf8_1.2.4                  Rsamtools_2.18.0           
 [25] rmarkdown_2.26              bit_4.0.5                   xfun_0.43                  
 [28] zlibbioc_1.48.2             cachem_1.0.8                progress_1.2.3             
 [31] blob_1.2.4                  DelayedArray_0.28.0         BiocParallel_1.36.0        
 [34] jpeg_0.1-10                 parallel_4.3.2              prettyunits_1.2.0          
 [37] cluster_2.1.6               VariantAnnotation_1.48.1    R6_2.5.1                   
 [40] stringi_1.8.3               RColorBrewer_1.1-3          rtracklayer_1.62.0         
 [43] pkgload_1.3.4               rpart_4.1.23                Rcpp_1.0.12                
 [46] SummarizedExperiment_1.32.0 knitr_1.45                  base64enc_0.1-3            
 [49] Matrix_1.6-5                nnet_7.3-19                 tidyselect_1.2.1           
 [52] rstudioapi_0.16.0           dichromat_2.0-0.1           abind_1.4-5                
 [55] yaml_2.3.8                  codetools_0.2-20            curl_5.2.1                 
 [58] lattice_0.22-6              tibble_3.2.1                Biobase_2.62.0             
 [61] KEGGREST_1.42.0             evaluate_0.23               foreign_0.8-86             
 [64] BiocFileCache_2.10.2        xml2_1.3.6                  Biostrings_2.70.3          
 [67] pillar_1.9.0                filelock_1.0.3              MatrixGenerics_1.14.0      
 [70] checkmate_2.3.1             generics_0.1.3              RCurl_1.98-1.14            
 [73] ensembldb_2.26.0            hms_1.1.3                   ggplot2_3.5.0              
 [76] munsell_0.5.1               scales_1.3.0                glue_1.7.0                 
 [79] lazyeval_0.2.2              Hmisc_5.1-2                 tools_4.3.2                
 [82] interp_1.1-6                BiocIO_1.12.0               data.table_1.15.4          
 [85] BSgenome_1.70.2             GenomicAlignments_1.38.2    XML_3.99-0.16.1            
 [88] latticeExtra_0.6-30         AnnotationDbi_1.64.1        colorspace_2.1-0           
 [91] GenomeInfoDbData_1.2.11     htmlTable_2.4.2             restfulr_0.0.15            
 [94] Formula_1.2-5               cli_3.6.2                   rappdirs_0.3.3             
 [97] fansi_1.0.6                 S4Arrays_1.2.1              dplyr_1.1.4                
[100] AnnotationFilter_1.26.0     gtable_0.3.4                digest_0.6.35              
[103] SparseArray_1.2.4           rjson_0.2.21                htmlwidgets_1.6.4          
[106] memoise_2.0.1               htmltools_0.5.8.1           lifecycle_1.0.4            
[109] httr_1.4.7                  bit64_4.0.5