Modeling continuous cell-level covariates
Collapse using mean value for pseudobulk data
Run on 2025-10-07 17:28:56
1 Introduction
Since read counts are summed across cells in a pseudobulk approach, modeling continuous cell-level covariates also requires a collapsing step. Here we summarize the values of a variable from a set of cells using the mean, and store the value for each cell type. Including these variables in a regression formula uses the summarized values from the corresponding cell type.
We demonstrate this feature on a lightly modified analysis of PBMCs from 8 individuals stimulated with interferon-β (Kang, et al, 2018, Nature Biotech).
2 Standard processing
Here is the code from the main vignette:
library(dreamlet)
library(muscat)
library(ExperimentHub)
library(scater)
# Download data, specifying EH2259 for the Kang, et al study
eh <- ExperimentHub()
sce <- eh[["EH2259"]]
# only keep singlet cells with sufficient reads
sce <- sce[rowSums(counts(sce) > 0) > 0, ]
sce <- sce[, colData(sce)$multiplets == "singlet"]
# compute QC metrics
qc <- perCellQCMetrics(sce)
# remove cells with few or many detected genes
ol <- isOutlier(metric = qc$detected, nmads = 2, log = TRUE)
sce <- sce[, !ol]
# set variable indicating stimulated (stim) or control (ctrl)
sce$StimStatus <- sce$stimIn many datasets, continuous cell-level variables could be mapped reads, gene count, mitochondrial rate, etc. There are no continuous cell-level variables in this dataset, so we can simulate two from a normal distribution:
sce$value1 <- rnorm(ncol(sce))
sce$value2 <- rnorm(ncol(sce))3 Pseudobulk
Now compute the pseudobulk using standard code:
sce$id <- paste0(sce$StimStatus, sce$ind)
# Create pseudobulk
pb <- aggregateToPseudoBulk(sce,
assay = "counts",
cluster_id = "cell",
sample_id = "id",
verbose = FALSE
)The means per variable, cell type, and sample are stored in the pseudobulk SingleCellExperiment object:
metadata(pb)$aggr_means## # A tibble: 128 × 5
## # Groups: cell [8]
## cell id cluster value1 value2
## <fct> <fct> <dbl> <dbl> <dbl>
## 1 B cells ctrl101 3.96 -0.0514 -0.00671
## 2 B cells ctrl1015 4.00 -0.00264 -0.0191
## 3 B cells ctrl1016 4 0.0950 -0.118
## 4 B cells ctrl1039 4.04 -0.129 -0.148
## 5 B cells ctrl107 4 0.128 0.0147
## 6 B cells ctrl1244 4 0.123 -0.162
## 7 B cells ctrl1256 4.01 0.0265 -0.0922
## 8 B cells ctrl1488 4.02 -0.0736 -0.00790
## 9 B cells stim101 4.09 -0.0852 0.0545
## 10 B cells stim1015 4.06 -0.0598 0.0387
## # ℹ 118 more rows4 Analysis
Including these variables in a regression formula uses the summarized values from the corresponding cell type. This happens behind the scenes, so the user doesn’t need to distinguish bewteen sample-level variables stored in colData(pb) and cell-level variables stored in metadata(pb)$aggr_means.
Variance partition and hypothesis testing proceeds as ususal:
form <- ~ StimStatus + value1 + value2
# Normalize and apply voom/voomWithDreamWeights
res.proc <- processAssays(pb, form, min.count = 5)
# run variance partitioning analysis
vp.lst <- fitVarPart(res.proc, form)
# Summarize variance fractions genome-wide for each cell type
plotVarPart(vp.lst, label.angle = 60)
# Differential expression analysis within each assay
res.dl <- dreamlet(res.proc, form)
# dreamlet results include coefficients for value1 and value2
res.dl## class: dreamletResult
## assays(8): B cells CD14+ Monocytes ... Megakaryocytes NK cells
## Genes:
## min: 164
## max: 5262
## details(7): assay n_retain ... n_errors error_initial
## coefNames(4): (Intercept) StimStatusstim value1 value25 Details
A variable in colData(sce) is handled according to if the variable is
- continuous: the mean per donor/cell type is stored in
metadata(pb)$aggr_means - discrete
- [constant within each donor/cell type] it is stored in
colData(pb) - [varies within each donor/cell type] there is no good way to summarize it. The variable is dropped.
- [constant within each donor/cell type] it is stored in
6 Session Info
## R version 4.5.1 Patched (2025-08-23 r88802)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.3 LTS
##
## Matrix products: default
## BLAS: /home/biocbuild/bbs-3.22-bioc/R/lib/libRblas.so
## LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.12.0 LAPACK version 3.12.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_GB LC_COLLATE=C
## [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
## [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## time zone: America/New_York
## tzcode source: system (glibc)
##
## attached base packages:
## [1] stats4 stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] muscData_1.23.0 scater_1.37.0
## [3] scuttle_1.19.0 ExperimentHub_2.99.5
## [5] AnnotationHub_3.99.6 BiocFileCache_2.99.6
## [7] dbplyr_2.5.1 muscat_1.23.1
## [9] dreamlet_1.7.1 SingleCellExperiment_1.31.1
## [11] SummarizedExperiment_1.39.2 Biobase_2.69.1
## [13] GenomicRanges_1.61.5 Seqinfo_0.99.2
## [15] IRanges_2.43.5 S4Vectors_0.47.4
## [17] BiocGenerics_0.55.1 generics_0.1.4
## [19] MatrixGenerics_1.21.0 matrixStats_1.5.0
## [21] variancePartition_1.39.3 BiocParallel_1.43.4
## [23] limma_3.65.5 ggplot2_4.0.0
## [25] BiocStyle_2.37.1
##
## loaded via a namespace (and not attached):
## [1] bitops_1.0-9 httr_1.4.7
## [3] RColorBrewer_1.1-3 doParallel_1.0.17
## [5] Rgraphviz_2.53.0 numDeriv_2016.8-1.1
## [7] sctransform_0.4.2 tools_4.5.1
## [9] backports_1.5.0 utf8_1.2.6
## [11] R6_2.6.1 metafor_4.8-0
## [13] mgcv_1.9-3 GetoptLong_1.0.5
## [15] withr_3.0.2 gridExtra_2.3
## [17] prettyunits_1.2.0 fdrtool_1.2.18
## [19] cli_3.6.5 sandwich_3.1-1
## [21] labeling_0.4.3 slam_0.1-55
## [23] sass_0.4.10 KEGGgraph_1.69.0
## [25] SQUAREM_2021.1 mvtnorm_1.3-3
## [27] S7_0.2.0 blme_1.0-6
## [29] mixsqp_0.3-54 zenith_1.11.0
## [31] dichromat_2.0-0.1 parallelly_1.45.1
## [33] invgamma_1.2 RSQLite_2.4.3
## [35] shape_1.4.6.1 gtools_3.9.5
## [37] dplyr_1.1.4 Matrix_1.7-4
## [39] metadat_1.4-0 ggbeeswarm_0.7.2
## [41] abind_1.4-8 lifecycle_1.0.4
## [43] multcomp_1.4-28 yaml_2.3.10
## [45] edgeR_4.7.5 mathjaxr_1.8-0
## [47] gplots_3.2.0 SparseArray_1.9.1
## [49] grid_4.5.1 blob_1.2.4
## [51] crayon_1.5.3 lattice_0.22-7
## [53] beachmat_2.25.5 msigdbr_25.1.1
## [55] annotate_1.87.0 KEGGREST_1.49.1
## [57] magick_2.9.0 pillar_1.11.1
## [59] knitr_1.50 ComplexHeatmap_2.25.2
## [61] rjson_0.2.23 boot_1.3-32
## [63] estimability_1.5.1 corpcor_1.6.10
## [65] future.apply_1.20.0 codetools_0.2-20
## [67] glue_1.8.0 data.table_1.17.8
## [69] vctrs_0.6.5 png_0.1-8
## [71] Rdpack_2.6.4 gtable_0.3.6
## [73] assertthat_0.2.1 cachem_1.1.0
## [75] zigg_0.0.2 xfun_0.53
## [77] rbibutils_2.3 S4Arrays_1.9.1
## [79] Rfast_2.1.5.1 coda_0.19-4.1
## [81] reformulas_0.4.1 survival_3.8-3
## [83] iterators_1.0.14 tinytex_0.57
## [85] statmod_1.5.0 TH.data_1.1-4
## [87] nlme_3.1-168 pbkrtest_0.5.5
## [89] bit64_4.6.0-1 filelock_1.0.3
## [91] progress_1.2.3 EnvStats_3.1.0
## [93] bslib_0.9.0 TMB_1.9.17
## [95] irlba_2.3.5.1 vipor_0.4.7
## [97] KernSmooth_2.23-26 colorspace_2.1-2
## [99] rmeta_3.0 DBI_1.2.3
## [101] DESeq2_1.49.4 tidyselect_1.2.1
## [103] emmeans_1.11.2-8 bit_4.6.0
## [105] compiler_4.5.1 curl_7.0.0
## [107] httr2_1.2.1 graph_1.87.0
## [109] BiocNeighbors_2.3.1 DelayedArray_0.35.3
## [111] bookdown_0.45 scales_1.4.0
## [113] caTools_1.18.3 remaCor_0.0.20
## [115] rappdirs_0.3.3 stringr_1.5.2
## [117] digest_0.6.37 minqa_1.2.8
## [119] rmarkdown_2.30 aod_1.3.3
## [121] XVector_0.49.1 RhpcBLASctl_0.23-42
## [123] htmltools_0.5.8.1 pkgconfig_2.0.3
## [125] lme4_1.1-37 sparseMatrixStats_1.21.0
## [127] lpsymphony_1.37.0 mashr_0.2.79
## [129] fastmap_1.2.0 rlang_1.1.6
## [131] GlobalOptions_0.1.2 DelayedMatrixStats_1.31.0
## [133] farver_2.1.2 jquerylib_0.1.4
## [135] IHW_1.37.0 zoo_1.8-14
## [137] jsonlite_2.0.0 BiocSingular_1.25.0
## [139] RCurl_1.98-1.17 magrittr_2.0.4
## [141] Rcpp_1.1.0 babelgene_22.9
## [143] viridis_0.6.5 EnrichmentBrowser_2.39.2
## [145] stringi_1.8.7 MASS_7.3-65
## [147] plyr_1.8.9 listenv_0.9.1
## [149] parallel_4.5.1 ggrepel_0.9.6
## [151] Biostrings_2.77.2 splines_4.5.1
## [153] hms_1.1.3 circlize_0.4.16
## [155] locfit_1.5-9.12 reshape2_1.4.4
## [157] ScaledMatrix_1.17.0 BiocVersion_3.22.0
## [159] XML_3.99-0.19 evaluate_1.0.5
## [161] RcppParallel_5.1.11-1 BiocManager_1.30.26
## [163] nloptr_2.2.1 foreach_1.5.2
## [165] tidyr_1.3.1 purrr_1.1.0
## [167] future_1.67.0 clue_0.3-66
## [169] scattermore_1.2 ashr_2.2-63
## [171] rsvd_1.0.5 broom_1.0.10
## [173] xtable_1.8-4 fANCOVA_0.6-1
## [175] viridisLite_0.4.2 truncnorm_1.0-9
## [177] tibble_3.3.0 lmerTest_3.1-3
## [179] glmmTMB_1.1.12 memoise_2.0.1
## [181] beeswarm_0.4.0 AnnotationDbi_1.71.1
## [183] cluster_2.1.8.1 globals_0.18.0
## [185] GSEABase_1.71.1