Modeling continuous cell-level covariates
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).
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:
Pseudobulk
Now compute the pseudobulk using standard code:
sce$id <- paste0(sce$StimStatus, sce$ind)
# Create pseudobulk
pb <- aggregateToPseudoBulk(sce,
assay = "counts",
cluster_id = "cell",
sample_id = "id",
verbose = FALSE
)The means per variable, cell type, and sample are stored in the
pseudobulk SingleCellExperiment object:
## # A tibble: 128 × 5
## # Groups: cell [8]
## cell id cluster value1 value2
## <fct> <fct> <dbl> <dbl> <dbl>
## 1 B cells ctrl101 3.96 0.119 -0.126
## 2 B cells ctrl1015 4.00 -0.0690 -0.0141
## 3 B cells ctrl1016 4 -0.154 -0.00902
## 4 B cells ctrl1039 4.04 0.00514 -0.104
## 5 B cells ctrl107 4 0.0544 0.135
## 6 B cells ctrl1244 4 0.0638 -0.0771
## 7 B cells ctrl1256 4.01 0.108 -0.0260
## 8 B cells ctrl1488 4.02 0.0333 0.112
## 9 B cells stim101 4.09 0.124 0.00736
## 10 B cells stim1015 4.06 0.0168 0.0751
## # ℹ 118 more rowsAnalysis
Including these variables in a regression formula uses the summarized
values from the corresponding cell type. This happens behind the scenes,
so the user doesn’t need to distinguish bewteen sample-level variables
stored in colData(pb) and cell-level variables stored in
metadata(pb)$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 value2Session Info
## R version 4.4.2 (2024-10-31)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.1 LTS
##
## Matrix products: default
## BLAS: /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3
## LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so; LAPACK version 3.12.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_US.UTF-8 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: Etc/UTC
## tzcode source: system (glibc)
##
## attached base packages:
## [1] stats4 stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] muscData_1.20.0 scater_1.35.0
## [3] scuttle_1.17.0 ExperimentHub_2.15.0
## [5] AnnotationHub_3.15.0 BiocFileCache_2.15.0
## [7] dbplyr_2.5.0 muscat_1.21.0
## [9] dreamlet_1.5.0 SingleCellExperiment_1.29.1
## [11] SummarizedExperiment_1.37.0 Biobase_2.67.0
## [13] GenomicRanges_1.59.0 GenomeInfoDb_1.43.1
## [15] IRanges_2.41.1 S4Vectors_0.45.2
## [17] BiocGenerics_0.53.3 generics_0.1.3
## [19] MatrixGenerics_1.19.0 matrixStats_1.4.1
## [21] variancePartition_1.37.1 BiocParallel_1.41.0
## [23] limma_3.63.2 ggplot2_3.5.1
## [25] BiocStyle_2.35.0
##
## 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_2.51.0 numDeriv_2016.8-1.1
## [7] tools_4.4.2 sctransform_0.4.1
## [9] backports_1.5.0 utf8_1.2.4
## [11] R6_2.5.1 metafor_4.6-0
## [13] mgcv_1.9-1 GetoptLong_1.0.5
## [15] withr_3.0.2 prettyunits_1.2.0
## [17] gridExtra_2.3 cli_3.6.3
## [19] labeling_0.4.3 sass_0.4.9
## [21] KEGGgraph_1.67.0 SQUAREM_2021.1
## [23] mvtnorm_1.3-2 blme_1.0-6
## [25] mixsqp_0.3-54 zenith_1.9.0
## [27] parallelly_1.39.0 invgamma_1.1
## [29] RSQLite_2.3.8 shape_1.4.6.1
## [31] gtools_3.9.5 dplyr_1.1.4
## [33] Matrix_1.7-1 metadat_1.2-0
## [35] ggbeeswarm_0.7.2 fansi_1.0.6
## [37] abind_1.4-8 lifecycle_1.0.4
## [39] yaml_2.3.10 edgeR_4.5.0
## [41] mathjaxr_1.6-0 gplots_3.2.0
## [43] SparseArray_1.7.2 grid_4.4.2
## [45] blob_1.2.4 crayon_1.5.3
## [47] lattice_0.22-6 beachmat_2.23.1
## [49] msigdbr_7.5.1 annotate_1.85.0
## [51] KEGGREST_1.47.0 sys_3.4.3
## [53] maketools_1.3.1 pillar_1.9.0
## [55] knitr_1.49 ComplexHeatmap_2.23.0
## [57] rjson_0.2.23 boot_1.3-31
## [59] corpcor_1.6.10 future.apply_1.11.3
## [61] codetools_0.2-20 glue_1.8.0
## [63] data.table_1.16.2 vctrs_0.6.5
## [65] png_0.1-8 Rdpack_2.6.2
## [67] gtable_0.3.6 assertthat_0.2.1
## [69] cachem_1.1.0 xfun_0.49
## [71] mime_0.12 rbibutils_2.3
## [73] S4Arrays_1.7.1 Rfast_2.1.0
## [75] reformulas_0.4.0 iterators_1.0.14
## [77] statmod_1.5.0 nlme_3.1-166
## [79] pbkrtest_0.5.3 bit64_4.5.2
## [81] filelock_1.0.3 progress_1.2.3
## [83] EnvStats_3.0.0 bslib_0.8.0
## [85] TMB_1.9.15 irlba_2.3.5.1
## [87] vipor_0.4.7 KernSmooth_2.23-24
## [89] colorspace_2.1-1 rmeta_3.0
## [91] DBI_1.2.3 DESeq2_1.47.1
## [93] tidyselect_1.2.1 curl_6.0.1
## [95] bit_4.5.0 compiler_4.4.2
## [97] graph_1.85.0 BiocNeighbors_2.1.0
## [99] DelayedArray_0.33.2 scales_1.3.0
## [101] caTools_1.18.3 remaCor_0.0.18
## [103] rappdirs_0.3.3 stringr_1.5.1
## [105] digest_0.6.37 minqa_1.2.8
## [107] rmarkdown_2.29 aod_1.3.3
## [109] XVector_0.47.0 RhpcBLASctl_0.23-42
## [111] htmltools_0.5.8.1 pkgconfig_2.0.3
## [113] lme4_1.1-35.5 sparseMatrixStats_1.19.0
## [115] mashr_0.2.79 fastmap_1.2.0
## [117] rlang_1.1.4 GlobalOptions_0.1.2
## [119] UCSC.utils_1.3.0 DelayedMatrixStats_1.29.0
## [121] farver_2.1.2 jquerylib_0.1.4
## [123] jsonlite_1.8.9 BiocSingular_1.23.0
## [125] RCurl_1.98-1.16 magrittr_2.0.3
## [127] GenomeInfoDbData_1.2.13 munsell_0.5.1
## [129] Rcpp_1.0.13-1 babelgene_22.9
## [131] viridis_0.6.5 EnrichmentBrowser_2.37.0
## [133] RcppZiggurat_0.1.6 stringi_1.8.4
## [135] zlibbioc_1.52.0 MASS_7.3-61
## [137] plyr_1.8.9 parallel_4.4.2
## [139] listenv_0.9.1 ggrepel_0.9.6
## [141] Biostrings_2.75.1 splines_4.4.2
## [143] hms_1.1.3 circlize_0.4.16
## [145] locfit_1.5-9.10 buildtools_1.0.0
## [147] reshape2_1.4.4 ScaledMatrix_1.15.0
## [149] BiocVersion_3.21.1 XML_3.99-0.17
## [151] evaluate_1.0.1 RcppParallel_5.1.9
## [153] BiocManager_1.30.25 nloptr_2.1.1
## [155] foreach_1.5.2 tidyr_1.3.1
## [157] purrr_1.0.2 future_1.34.0
## [159] clue_0.3-66 scattermore_1.2
## [161] ashr_2.2-63 rsvd_1.0.5
## [163] broom_1.0.7 xtable_1.8-4
## [165] fANCOVA_0.6-1 viridisLite_0.4.2
## [167] truncnorm_1.0-9 tibble_3.2.1
## [169] lmerTest_3.1-3 glmmTMB_1.1.10
## [171] memoise_2.0.1 beeswarm_0.4.0
## [173] AnnotationDbi_1.69.0 cluster_2.1.6
## [175] globals_0.16.3 GSEABase_1.69.0