sample_frac function changes the cell order

[aodainic7]

Hey,
when I try subsetting a large Seurat object to reduce the computing time, the sample_frac() function changes the cell order, so that the Seurat functions do not work anymore. To repeat the error try the code:

pbmc_small = SeuratObject::pbmc_small

pbmc_small_subset <- pbmc_small |> sample_frac(0.9)

pbmc_small_subset <- RunPCA(pbmc_small_subset, reduction.name = 'pca', assay = "RNA")

The error I'm getting is :
Error in validObject(object = x) : invalid class “Seurat” object: 1: all cells in assays must be in the same order as the Seurat object invalid class “Seurat” object: 2: all cells in reductions must be in the same order as the Seurat object invalid class “Seurat” object: 3: all cells in graphs must be in the same order as the Seurat object (offending: RNA_snn) invalid class “Seurat” object: 4: 'active.idents' must be named with cell names

Thanks!

sessionInfo()
R version 4.2.0 (2022-04-22)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8 LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8 LC_PAPER=en_US.UTF-8
[8] LC_NAME=C LC_ADDRESS=C LC_TELEPHONE=C LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

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

other attached packages:
[1] cbmc.SeuratData_3.1.4 SeuratData_0.2.2 scclusteval_0.0.0.9000 SingleCellExperiment_1.18.1 SummarizedExperiment_1.26.1 Biobase_2.56.0 GenomicRanges_1.48.0
[8] GenomeInfoDb_1.32.4 IRanges_2.30.1 S4Vectors_0.34.0 BiocGenerics_0.42.0 MatrixGenerics_1.8.1 matrixStats_0.63.0 tidyseurat_0.5.9
[15] ttservice_0.2.2 RColorBrewer_1.1-3 patchwork_1.1.2 Seurat_4.9.9.9042 SeuratObject_4.9.9.9084 sp_1.6-0 lubridate_1.9.2
[22] forcats_1.0.0 stringr_1.5.0 dplyr_1.1.2 purrr_1.0.1 readr_2.1.4 tidyr_1.3.0 tibble_3.2.1
[29] ggplot2_3.4.2 tidyverse_2.0.0

loaded via a namespace (and not attached):
[1] spam_2.9-1 systemfonts_1.0.4 plyr_1.8.8 igraph_1.4.2 lazyeval_0.2.2 splines_4.2.0 RcppHNSW_0.4.1 listenv_0.9.0
[9] scattermore_0.8 digest_0.6.31 htmltools_0.5.5 fansi_1.0.4 magrittr_2.0.3 tensor_1.5 cluster_2.1.4 ROCR_1.0-11
[17] limma_3.52.4 tzdb_0.3.0 globals_0.16.2 timechange_0.2.0 spatstat.sparse_3.0-1 colorspace_2.1-0 rappdirs_0.3.3 ggrepel_0.9.3
[25] textshaping_0.3.6 xfun_0.39 crayon_1.5.2 RCurl_1.98-1.12 jsonlite_1.8.4 progressr_0.13.0 spatstat.data_3.0-1 survival_3.3-1
[33] zoo_1.8-12 glue_1.6.2 polyclip_1.10-4 gtable_0.3.3 zlibbioc_1.42.0 XVector_0.36.0 leiden_0.4.3 DelayedArray_0.22.0
[41] future.apply_1.10.0 abind_1.4-5 scales_1.2.1 spatstat.random_3.1-4 miniUI_0.1.1.1 Rcpp_1.0.10 viridisLite_0.4.1 xtable_1.8-4
[49] reticulate_1.28 dotCall64_1.0-2 htmlwidgets_1.6.2 httr_1.4.5 ellipsis_0.3.2 ica_1.0-3 farver_2.1.1 pkgconfig_2.0.3
[57] sass_0.4.5 uwot_0.1.14 deldir_1.0-6 utf8_1.2.3 here_1.0.1 labeling_0.4.2 tidyselect_1.2.0 rlang_1.1.0
[65] reshape2_1.4.4 later_1.3.0 cachem_1.0.7 munsell_0.5.0 tools_4.2.0 cli_3.6.1 generics_0.1.3 ggridges_0.5.4
[73] evaluate_0.20 fastmap_1.1.1 yaml_2.3.7 ragg_1.2.5 goftest_1.2-3 knitr_1.42 fitdistrplus_1.1-11 RANN_2.6.1
[81] pbapply_1.7-0 future_1.32.0 nlme_3.1-160 mime_0.12 compiler_4.2.0 rstudioapi_0.14 plotly_4.10.1 png_0.1-8
[89] spatstat.utils_3.0-2 bslib_0.4.2 stringi_1.7.12 RSpectra_0.16-1 lattice_0.20-45 Matrix_1.5-1 vctrs_0.6.2 pillar_1.9.0
[97] lifecycle_1.0.3 BiocManager_1.30.20 jquerylib_0.1.4 spatstat.geom_3.1-0 lmtest_0.9-40 RcppAnnoy_0.0.20 data.table_1.14.8 cowplot_1.1.1
[105] bitops_1.0-7 irlba_2.3.5.1 httpuv_1.6.9 R6_2.5.1 promises_1.2.0.1 renv_0.17.3 KernSmooth_2.23-20 gridExtra_2.3
[113] parallelly_1.35.0 codetools_0.2-18 fastDummies_1.6.3 MASS_7.3-58.1 rprojroot_2.0.3 withr_2.5.0 sctransform_0.3.5 GenomeInfoDbData_1.2.8
[121] parallel_4.2.0 hms_1.1.3 grid_4.2.0 rmarkdown_2.21 Rtsne_0.16 spatstat.explore_3.1-0 shiny_1.7.4

[william-hutchison]

Hello,

I believe changing the row order is the intended behaviour of sample_fraq() in dplyr, so it makes sense for the function to change cell order in tidyseurat.

You can see how sample_fraq() changes the row order in this example: https://dplyr.tidyverse.org/reference/sample_n.html

[stemangiola]

Yes it randomizes cells, but it should not break the object, I think cells are randomised but not for all assays or slots somehow.

You can see the seurat function here that does that for reference

subset <- SubsetData(object, max.cells.per.ident = n.cells, random.seed = NULL)

from

satijalab/seurat#243

possibly related

satijalab/seurat#5329