Chapter 6 Muraro human pancreas (CEL-seq)
6.1 Introduction
This performs an analysis of the Muraro et al. (2016) CEL-seq dataset, consisting of human pancreas cells from various donors.
6.2 Data loading
Converting back to Ensembl identifiers.
library(AnnotationHub)
edb <- AnnotationHub()[["AH73881"]]
gene.symb <- sub("__chr.*$", "", rownames(sce.muraro))
gene.ids <- mapIds(edb, keys=gene.symb,
keytype="SYMBOL", column="GENEID")
# Removing duplicated genes or genes without Ensembl IDs.
keep <- !is.na(gene.ids) & !duplicated(gene.ids)
sce.muraro <- sce.muraro[keep,]
rownames(sce.muraro) <- gene.ids[keep]6.3 Quality control
This dataset lacks mitochondrial genes so we will do without. For the one batch that seems to have a high proportion of low-quality cells, we compute an appropriate filter threshold using a shared median and MAD from the other batches (Figure 6.1).
library(scater)
stats <- perCellQCMetrics(sce.muraro)
qc <- quickPerCellQC(stats, percent_subsets="altexps_ERCC_percent",
batch=sce.muraro$donor, subset=sce.muraro$donor!="D28")
sce.muraro <- sce.muraro[,!qc$discard]colData(unfiltered) <- cbind(colData(unfiltered), stats)
unfiltered$discard <- qc$discard
gridExtra::grid.arrange(
plotColData(unfiltered, x="donor", y="sum", colour_by="discard") +
scale_y_log10() + ggtitle("Total count"),
plotColData(unfiltered, x="donor", y="detected", colour_by="discard") +
scale_y_log10() + ggtitle("Detected features"),
plotColData(unfiltered, x="donor", y="altexps_ERCC_percent",
colour_by="discard") + ggtitle("ERCC percent"),
ncol=2
)
Figure 6.1: Distribution of each QC metric across cells from each donor in the Muraro pancreas dataset. Each point represents a cell and is colored according to whether that cell was discarded.
We have a look at the causes of removal:
## low_lib_size low_n_features high_altexps_ERCC_percent
## 663 700 738
## discard
## 7736.4 Normalization
library(scran)
set.seed(1000)
clusters <- quickCluster(sce.muraro)
sce.muraro <- computeSumFactors(sce.muraro, clusters=clusters)
sce.muraro <- logNormCounts(sce.muraro)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.088 0.541 0.821 1.000 1.211 13.987plot(librarySizeFactors(sce.muraro), sizeFactors(sce.muraro), pch=16,
xlab="Library size factors", ylab="Deconvolution factors", log="xy")
Figure 6.2: Relationship between the library size factors and the deconvolution size factors in the Muraro pancreas dataset.
6.5 Variance modelling
We block on a combined plate and donor factor.
block <- paste0(sce.muraro$plate, "_", sce.muraro$donor)
dec.muraro <- modelGeneVarWithSpikes(sce.muraro, "ERCC", block=block)
top.muraro <- getTopHVGs(dec.muraro, prop=0.1)par(mfrow=c(8,4))
blocked.stats <- dec.muraro$per.block
for (i in colnames(blocked.stats)) {
current <- blocked.stats[[i]]
plot(current$mean, current$total, main=i, pch=16, cex=0.5,
xlab="Mean of log-expression", ylab="Variance of log-expression")
curfit <- metadata(current)
points(curfit$mean, curfit$var, col="red", pch=16)
curve(curfit$trend(x), col='dodgerblue', add=TRUE, lwd=2)
}
Figure 6.3: Per-gene variance as a function of the mean for the log-expression values in the Muraro pancreas dataset. Each point represents a gene (black) with the mean-variance trend (blue) fitted to the spike-in transcripts (red) separately for each donor.
6.6 Data integration
library(batchelor)
set.seed(1001010)
merged.muraro <- fastMNN(sce.muraro, subset.row=top.muraro,
batch=sce.muraro$donor)We use the proportion of variance lost as a diagnostic measure:
## D28 D29 D30 D31
## [1,] 0.060847 0.024121 0.000000 0.00000
## [2,] 0.002646 0.003018 0.062421 0.00000
## [3,] 0.003449 0.002641 0.002598 0.081626.8 Clustering
snn.gr <- buildSNNGraph(merged.muraro, use.dimred="corrected")
colLabels(merged.muraro) <- factor(igraph::cluster_walktrap(snn.gr)$membership)tab <- table(Cluster=colLabels(merged.muraro), CellType=sce.muraro$label)
library(pheatmap)
pheatmap(log10(tab+10), color=viridis::viridis(100))
Figure 6.4: Heatmap of the frequency of cells from each cell type label in each cluster.
## Donor
## Cluster D28 D29 D30 D31
## 1 104 6 57 112
## 2 59 21 77 97
## 3 12 75 64 43
## 4 28 149 126 120
## 5 87 261 277 214
## 6 21 7 54 26
## 7 1 6 6 37
## 8 6 6 5 2
## 9 11 68 5 30
## 10 4 2 5 8gridExtra::grid.arrange(
plotTSNE(merged.muraro, colour_by="label"),
plotTSNE(merged.muraro, colour_by="batch"),
ncol=2
)
Figure 6.5: Obligatory \(t\)-SNE plots of the Muraro pancreas dataset. Each point represents a cell that is colored by cluster (left) or batch (right).
Session Info
R Under development (unstable) (2024-10-21 r87258)
Platform: x86_64-pc-linux-gnu
Running under: Ubuntu 24.04.1 LTS
Matrix products: default
BLAS: /home/biocbuild/bbs-3.21-bioc/R/lib/libRblas.so
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.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] pheatmap_1.0.12 batchelor_1.23.0
[3] scran_1.35.0 scater_1.35.0
[5] ggplot2_3.5.1 scuttle_1.17.0
[7] ensembldb_2.31.0 AnnotationFilter_1.31.0
[9] GenomicFeatures_1.59.1 AnnotationDbi_1.69.0
[11] AnnotationHub_3.15.0 BiocFileCache_2.15.0
[13] dbplyr_2.5.0 scRNAseq_2.21.0
[15] SingleCellExperiment_1.29.1 SummarizedExperiment_1.37.0
[17] Biobase_2.67.0 GenomicRanges_1.59.1
[19] GenomeInfoDb_1.43.2 IRanges_2.41.2
[21] S4Vectors_0.45.2 BiocGenerics_0.53.3
[23] generics_0.1.3 MatrixGenerics_1.19.1
[25] matrixStats_1.5.0 BiocStyle_2.35.0
[27] rebook_1.17.0
loaded via a namespace (and not attached):
[1] RColorBrewer_1.1-3 jsonlite_1.8.9
[3] CodeDepends_0.6.6 magrittr_2.0.3
[5] ggbeeswarm_0.7.2 gypsum_1.3.0
[7] farver_2.1.2 rmarkdown_2.29
[9] BiocIO_1.17.1 vctrs_0.6.5
[11] DelayedMatrixStats_1.29.1 memoise_2.0.1
[13] Rsamtools_2.23.1 RCurl_1.98-1.16
[15] htmltools_0.5.8.1 S4Arrays_1.7.1
[17] curl_6.1.0 BiocNeighbors_2.1.2
[19] Rhdf5lib_1.29.0 SparseArray_1.7.2
[21] rhdf5_2.51.2 sass_0.4.9
[23] alabaster.base_1.7.2 bslib_0.8.0
[25] alabaster.sce_1.7.0 httr2_1.0.7
[27] cachem_1.1.0 ResidualMatrix_1.17.0
[29] GenomicAlignments_1.43.0 igraph_2.1.3
[31] mime_0.12 lifecycle_1.0.4
[33] pkgconfig_2.0.3 rsvd_1.0.5
[35] Matrix_1.7-1 R6_2.5.1
[37] fastmap_1.2.0 GenomeInfoDbData_1.2.13
[39] digest_0.6.37 colorspace_2.1-1
[41] dqrng_0.4.1 irlba_2.3.5.1
[43] ExperimentHub_2.15.0 RSQLite_2.3.9
[45] beachmat_2.23.6 labeling_0.4.3
[47] filelock_1.0.3 httr_1.4.7
[49] abind_1.4-8 compiler_4.5.0
[51] bit64_4.5.2 withr_3.0.2
[53] BiocParallel_1.41.0 viridis_0.6.5
[55] DBI_1.2.3 HDF5Array_1.35.3
[57] alabaster.ranges_1.7.0 alabaster.schemas_1.7.0
[59] rappdirs_0.3.3 DelayedArray_0.33.3
[61] bluster_1.17.0 rjson_0.2.23
[63] tools_4.5.0 vipor_0.4.7
[65] beeswarm_0.4.0 glue_1.8.0
[67] restfulr_0.0.15 rhdf5filters_1.19.0
[69] grid_4.5.0 Rtsne_0.17
[71] cluster_2.1.8 gtable_0.3.6
[73] metapod_1.15.0 BiocSingular_1.23.0
[75] ScaledMatrix_1.15.0 XVector_0.47.2
[77] ggrepel_0.9.6 BiocVersion_3.21.1
[79] pillar_1.10.1 limma_3.63.3
[81] dplyr_1.1.4 lattice_0.22-6
[83] rtracklayer_1.67.0 bit_4.5.0.1
[85] tidyselect_1.2.1 locfit_1.5-9.10
[87] Biostrings_2.75.3 knitr_1.49
[89] gridExtra_2.3 bookdown_0.42
[91] ProtGenerics_1.39.1 edgeR_4.5.1
[93] xfun_0.50 statmod_1.5.0
[95] UCSC.utils_1.3.0 lazyeval_0.2.2
[97] yaml_2.3.10 evaluate_1.0.3
[99] codetools_0.2-20 tibble_3.2.1
[101] alabaster.matrix_1.7.4 BiocManager_1.30.25
[103] graph_1.85.1 cli_3.6.3
[105] munsell_0.5.1 jquerylib_0.1.4
[107] Rcpp_1.0.14 dir.expiry_1.15.0
[109] png_0.1-8 XML_3.99-0.18
[111] parallel_4.5.0 blob_1.2.4
[113] sparseMatrixStats_1.19.0 bitops_1.0-9
[115] viridisLite_0.4.2 alabaster.se_1.7.0
[117] scales_1.3.0 purrr_1.0.2
[119] crayon_1.5.3 rlang_1.1.4
[121] cowplot_1.1.3 KEGGREST_1.47.0 References
Muraro, M. J., G. Dharmadhikari, D. Grun, N. Groen, T. Dielen, E. Jansen, L. van Gurp, et al. 2016. “A Single-Cell Transcriptome Atlas of the Human Pancreas.” Cell Syst 3 (4): 385–94.