single cell 10x single-cell analysis - part5
UC Davis Bioinformatics Core
Load the Seurat object
load(file="pca_sample_corrected.RData")
experiment.aggregate## Loading required package: Seurat## Loading required package: ggplot2## Loading required package: cowplot##
## Attaching package: 'cowplot'## The following object is masked from 'package:ggplot2':
##
## ggsave## Loading required package: Matrix## An object of class seurat in project scRNA workshop
## 11454 genes across 21288 samples.Identifying clusters
Seurat implements an graph-based clustering approach. Distances between the cells are calculated based on previously identified PCs. Seurat approach was heavily inspired by recent manuscripts which applied graph-based clustering approaches to scRNAseq data. Briefly, Seurat identify clusters of cells by a shared nearest neighbor (SNN) modularity optimization based clustering algorithm. First calculate k-nearest neighbors (KNN) and construct the SNN graph. Then optimize the modularity function to determine clusters. For a full description of the algorithms, see Waltman and van Eck (2013) The European Physical Journal B.
The FindClusters function implements the procedure, and contains a resolution parameter that sets the granularity of the downstream clustering, with increased values leading to a greater number of clusters. I tend to like to perform a series of resolutions, investigate and choose.
WARNING: TAKES A LONG TIME TO RUN
use.pcs = 1:35
experiment.aggregate <- FindClusters(
object = experiment.aggregate,
reduction.type = "pca",
dims.use = use.pcs,
resolution = seq(0.5,4,0.5),
print.output = FALSE,
save.SNN = TRUE
)
PrintFindClustersParams(object = experiment.aggregate)## Parameters used in latest FindClusters calculation run on: 2018-03-22 08:56:57
## =============================================================================
## Resolution: 0.5
## -----------------------------------------------------------------------------
## Modularity Function Algorithm n.start n.iter
## 1 1 100 10
## -----------------------------------------------------------------------------
## Reduction used k.param k.scale prune.SNN
## pca 30 25 0.0667
## -----------------------------------------------------------------------------
## Dims used in calculation
## =============================================================================
## 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
## 30 31 32 33 34 35Lets first investigate how many clusters each resolution produces and set it to the smallest resolutions of 0.5 (fewest clusters). finnaly lets produce a table of cluster to sample assignments.
sapply(grep("^res",colnames(experiment.aggregate@meta.data),value = TRUE),
function(x) length(unique(experiment.aggregate@meta.data[,x])))## res.0.5 res.1 res.1.5 res.2 res.2.5 res.3 res.3.5 res.4
## 16 23 30 37 41 43 46 50experiment.aggregate <- SetAllIdent(experiment.aggregate, id = "res.0.5")
table(experiment.aggregate@ident,experiment.aggregate@meta.data$orig.ident)##
## sample1 sample2 sample3
## 0 1084 1138 523
## 1 799 974 954
## 2 760 724 747
## 3 727 656 591
## 4 632 649 560
## 5 484 626 579
## 6 464 491 579
## 7 315 441 375
## 8 338 399 304
## 9 248 318 326
## 10 183 332 257
## 11 204 261 263
## 12 176 233 204
## 13 213 182 216
## 14 123 201 132
## 15 93 147 63tSNE dimensionality reduction plots are then used to visualise clustering results. As input to the tSNE, you should use the same PCs as input to the clustering analysis.
experiment.aggregate <- RunTSNE(
object = experiment.aggregate,
reduction.use = "pca",
dims.use = use.pcs,
do.fast = TRUE)Plot TSNE coloring by the slot ‘ident’ (default).
TSNEPlot(object = experiment.aggregate, pt.size=0.5)
Plot TSNE coloring by the slot ‘orig.ident’ (sample names).
TSNEPlot(object = experiment.aggregate, group.by="orig.ident", pt.size=0.5)
Plot TSNE coloring by the clustering resolution 4
TSNEPlot(object = experiment.aggregate, group.by="res.4", pt.size=0.5, do.label = TRUE)
FeaturePlot can be used to color cells with a ‘feature’, non categorical data, like number of UMIs
FeaturePlot(experiment.aggregate, features.plot=c('nUMI'), pt.size=0.5)
and number of genes present
FeaturePlot(experiment.aggregate, features.plot=c('nGene'), pt.size=0.5)
and percent mitochondrial
FeaturePlot(experiment.aggregate, features.plot=c('percent.mito'), pt.size=0.5)
Building a tree relating the ‘average’ cell from each cluster. Tree is estimated based on a distance matrix constructed in either gene expression space or PCA space.
experiment.aggregate <- BuildClusterTree(
experiment.aggregate,
pcs.use = use.pcs,
do.reorder = F,
reorder.numeric = F,
do.plot=F)## [1] "Finished averaging RNA for cluster 0"
## [1] "Finished averaging RNA for cluster 1"
## [1] "Finished averaging RNA for cluster 2"
## [1] "Finished averaging RNA for cluster 3"
## [1] "Finished averaging RNA for cluster 4"
## [1] "Finished averaging RNA for cluster 5"
## [1] "Finished averaging RNA for cluster 6"
## [1] "Finished averaging RNA for cluster 7"
## [1] "Finished averaging RNA for cluster 8"
## [1] "Finished averaging RNA for cluster 9"
## [1] "Finished averaging RNA for cluster 10"
## [1] "Finished averaging RNA for cluster 11"
## [1] "Finished averaging RNA for cluster 12"
## [1] "Finished averaging RNA for cluster 13"
## [1] "Finished averaging RNA for cluster 14"
## [1] "Finished averaging RNA for cluster 15"PlotClusterTree(experiment.aggregate)
Plot the split at node 30
ColorTSNESplit(experiment.aggregate, node = 30)
TSNEPlot(object = experiment.aggregate, pt.size=0.5, do.label = TRUE)
experiment.merged <- RenameIdent(
object = experiment.aggregate,
old.ident.name = c('14'),
new.ident.name = '0'
)
TSNEPlot(object = experiment.merged, pt.size=0.5, do.label = T)
Identifying Marker Genes
Seurat can help you find markers that define clusters via differential expression.
FindMarkers identifies markers for a cluster relative to all other clusters.
FindAllMarkers does so for all clusters
FindAllMarkersNode defines all markers that split a Node (Warning: need to validate)
?FindMarkers
markers = FindMarkers(experiment.aggregate, ident.1=c(0,14), genes.use=rownames(experiment.aggregate@scale.data))
head(markers)## p_val avg_logFC pct.1 pct.2 p_val_adj
## Ctxn3 0 1.665902 0.923 0.143 0
## Txn1 0 1.559953 0.978 0.663 0
## Ly86 0 1.490500 0.631 0.092 0
## Lpar3 0 1.451481 0.759 0.015 0
## Gm765 0 1.445594 0.818 0.059 0
## Cd55 0 1.336349 0.931 0.261 0dim(markers)## [1] 451 5table(markers$avg_logFC > 0)##
## FALSE TRUE
## 364 87pct.1 and pct.2 are the proportion of cells with expression above 0 in ident.1 and ident.2 respectively. p_val is the raw p_value associated with the differntial expression test with adjusted value in p_val_adj. avg_logFC is the average log fold change difference between the two groups.
avg_diff (lines 130, 193 and) appears to be the difference in log(x = mean(x = exp(x = x) - 1) + 1) between groups. It doesn’t seem like this should work out to be the signed ratio of pct.1 to pct.2 so I must be missing something. It doesn’t seem to be related at all to how the p-values are calculated so maybe it doesn’t matter so much, and the sign is probably going to be pretty robust to how expression is measured.
Can use a violin plot to visualize the expression pattern of some markers
VlnPlot(object = experiment.aggregate, features.plot = rownames(markers)[1:2])
Or a feature plot
FeaturePlot(
experiment.aggregate,
head(rownames(markers)),
cols.use = c("lightgrey", "blue"),
nCol = 3
)
FeaturePlot(
experiment.aggregate,
"Lpar3",
cols.use = c("lightgrey", "blue")
)
FindAllMarkers can be used to automate the process across all genes. WARNING: TAKES A LONG TIME TO RUN
markers_all <- FindAllMarkers(
object = experiment.aggregate,
only.pos = TRUE,
min.pct = 0.25,
thresh.use = 0.25
)
dim(markers_all)## [1] 7271 7head(markers_all)## p_val avg_logFC pct.1 pct.2 p_val_adj cluster gene
## Ctxn3 0 1.578061 0.933 0.161 0 0 Ctxn3
## Txn1 0 1.570280 0.988 0.669 0 0 Txn1
## Lpar3 0 1.529082 0.841 0.022 0 0 Lpar3
## Ly86 0 1.513758 0.673 0.099 0 0 Ly86
## Gm765 0 1.472250 0.861 0.071 0 0 Gm765
## Cd55 0 1.330898 0.950 0.274 0 0 Cd55table(table(markers_all$gene))##
## 1 2 3 4 5 6 7 9
## 1182 908 560 327 173 51 15 1markers_all_single <- markers_all[markers_all$gene %in% names(table(markers_all$gene))[table(markers_all$gene) == 1],]
dim(markers_all_single)## [1] 1182 7table(table(markers_all_single$gene))##
## 1
## 1182table(markers_all_single$cluster)##
## 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
## 84 21 90 110 248 28 29 43 16 45 194 133 62 58 11 10head(markers_all_single)## p_val avg_logFC pct.1 pct.2 p_val_adj cluster gene
## Lpar3 0 1.5290822 0.841 0.022 0 0 Lpar3
## 9130204L05Rik 0 0.7503062 0.303 0.013 0 0 9130204L05Rik
## Ms4a3 0 0.5940921 0.402 0.021 0 0 Ms4a3
## Barx2 0 0.5135031 0.326 0.007 0 0 Barx2
## Prkcq 0 0.4908947 0.332 0.009 0 0 Prkcq
## Zdhhc12 0 0.4667916 0.412 0.068 0 0 Zdhhc12Plot a heatmap of genes by cluster for the top 5 marker genes per cluster
#install.packages("dplyr")
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, uniontop5 <- markers_all_single %>% group_by(cluster) %>% top_n(5, avg_logFC)
dim(top5)## [1] 80 7DoHeatmap(
object = experiment.aggregate,
genes.use = top5$gene,
slim.col.label = TRUE,
remove.key = TRUE
)
# Get expression of
getGeneClusterMeans <- function(gene, cluster){
x <- experiment.aggregate@data[gene,]
m <- tapply(x, ifelse(experiment.aggregate@ident == cluster, 1, 0), mean)
mean.in.cluster <- m[2]
mean.out.of.cluster <- m[1]
return(list(mean.in.cluster = mean.in.cluster, mean.out.of.cluster = mean.out.of.cluster))
}
## for sake of time only using first six (head)
means <- mapply(getGeneClusterMeans, head(markers_all[,"gene"]), head(markers_all[,"cluster"]))
means <- matrix(unlist(means), ncol = 2, byrow = T)
colnames(means) <- c("mean.in.cluster", "mean.out.of.cluster")
markers_all2 <- cbind(head(markers_all), means)
head(markers_all2)## p_val avg_logFC pct.1 pct.2 p_val_adj cluster gene mean.in.cluster
## Ctxn3 0 1.578061 0.933 0.161 0 0 Ctxn3 2.011848
## Txn1 0 1.570280 0.988 0.669 0 0 Txn1 3.532586
## Lpar3 0 1.529082 0.841 0.022 0 0 Lpar3 1.361024
## Ly86 0 1.513758 0.673 0.099 0 0 Ly86 1.318193
## Gm765 0 1.472250 0.861 0.071 0 0 Gm765 1.472922
## Cd55 0 1.330898 0.950 0.274 0 0 Cd55 1.915879
## mean.out.of.cluster
## Ctxn3 0.27084140
## Txn1 1.44303974
## Lpar3 0.02792977
## Ly86 0.14667298
## Gm765 0.10476314
## Cd55 0.41644465Finishing up clusters.
At this point in time you should use the tree, markers, domain knowledge, and goals to finalize your clusters. This may mean adjusting PCA to use, mergers clusters together, choosing a new resolutions, etc. When finished you can further name it cluster by something more informative. Ex.
experiment.clusters <- experiment.aggregate
experiment.clusters <- RenameIdent(
object = experiment.clusters,
old.ident.name = c('0'),
new.ident.name = 'cell_type_A'
)
TSNEPlot(object = experiment.clusters, pt.size=0.5, do.label = T)
experiment.aggregate <- AddMetaData(
object = experiment.aggregate,
metadata = experiment.aggregate@ident,
col.name = "finalcluster")
head(experiment.aggregate@meta.data)## nGene nUMI orig.ident percent.mito batchid
## AAAACCTGAGATCACGG-sample1 1309 2435 sample1 0.075154004 UCD_VitE_Def
## AAAACCTGAGCATCATC-sample1 1331 2650 sample1 0.008301887 UCD_VitE_Def
## AAAACCTGAGCGCTCCA-sample1 1507 2817 sample1 0.038693646 UCD_VitE_Def
## AAAACCTGAGTGGGATC-sample1 1891 3625 sample1 0.005517241 UCD_VitE_Def
## AAAACCTGCAGACAGGT-sample1 786 1091 sample1 0.013748854 UCD_VitE_Def
## AAAACCTGGTATAGGTA-sample1 2505 7248 sample1 0.027179912 UCD_VitE_Def
## res.0.5 res.1 res.1.5 res.2 res.2.5 res.3
## AAAACCTGAGATCACGG-sample1 2 5 2 1 1 0
## AAAACCTGAGCATCATC-sample1 0 1 5 3 18 17
## AAAACCTGAGCGCTCCA-sample1 2 5 2 1 1 0
## AAAACCTGAGTGGGATC-sample1 9 8 9 24 24 24
## AAAACCTGCAGACAGGT-sample1 3 2 8 6 7 6
## AAAACCTGGTATAGGTA-sample1 12 12 13 12 11 10
## res.3.5 res.4 finalcluster
## AAAACCTGAGATCACGG-sample1 0 0 2
## AAAACCTGAGCATCATC-sample1 16 12 0
## AAAACCTGAGCGCTCCA-sample1 0 0 2
## AAAACCTGAGTGGGATC-sample1 25 24 9
## AAAACCTGCAGACAGGT-sample1 21 13 3
## AAAACCTGGTATAGGTA-sample1 24 23 12Subsetting samples
experiment.sample2 <- SubsetData(
object = experiment.aggregate,
cells.use = rownames(experiment.aggregate@meta.data)[experiment.aggregate@meta.data$orig.ident %in% c("sample2")])
TSNEPlot(object = experiment.sample2, group.by="res.0.5", pt.size=0.5, do.label = TRUE)
FeaturePlot(experiment.sample2, features.plot=c('Calca'), pt.size=0.5)
FeaturePlot(experiment.sample2, features.plot=c('Adcyap1'), pt.size=0.5)
Adding in a new metadata column representing samples within clusters
samplecluster = paste(experiment.aggregate@meta.data$orig.ident,experiment.aggregate@meta.data$finalcluster,sep = '-')
names(samplecluster) = rownames(experiment.aggregate@meta.data)
head(samplecluster)## AAAACCTGAGATCACGG-sample1 AAAACCTGAGCATCATC-sample1
## "sample1-2" "sample1-0"
## AAAACCTGAGCGCTCCA-sample1 AAAACCTGAGTGGGATC-sample1
## "sample1-2" "sample1-9"
## AAAACCTGCAGACAGGT-sample1 AAAACCTGGTATAGGTA-sample1
## "sample1-3" "sample1-12"experiment.aggregate <- AddMetaData(
object = experiment.aggregate,
metadata = samplecluster,
col.name = "samplecluster")
# set the identity to the new variable
experiment.aggregate <- SetAllIdent(experiment.aggregate, id = "samplecluster")
markers.comp <- FindMarkers(experiment.aggregate, ident.1 = "sample1-0", ident.2= c("sample2-0","sample3-0"))
markers.comp## p_val avg_logFC pct.1 pct.2 p_val_adj
## 9130204L05Rik 1.244451e-20 0.3814872 0.398 0.241 1.425394e-16
## Apoe 2.310683e-03 -0.2991319 0.267 0.313 1.000000e+00DoHeatmap(experiment.aggregate,
use.scaled = TRUE,
cells.use=rownames(experiment.aggregate@meta.data)[experiment.aggregate@meta.data$samplecluster %in% c( "sample1-0", "sample2-0" )],
genes.use = rownames(markers.comp),
slim.col.label = TRUE
)
experiment.aggregate <- SetAllIdent(experiment.aggregate, id = "finalcluster")And last lets save all the objects in our session.
save(list=ls(), file="clusters_seurat_object.RData")Session Information
sessionInfo()## R version 3.4.4 (2018-03-15)
## Platform: x86_64-apple-darwin15.6.0 (64-bit)
## Running under: macOS High Sierra 10.13.3
##
## Matrix products: default
## BLAS: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRblas.0.dylib
## LAPACK: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRlapack.dylib
##
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] dplyr_0.7.4 bindrcpp_0.2 Seurat_2.2.1 Matrix_1.2-12 cowplot_0.9.2
## [6] ggplot2_2.2.1
##
## loaded via a namespace (and not attached):
## [1] diffusionMap_1.1-0 Rtsne_0.13 VGAM_1.0-5
## [4] colorspace_1.3-2 ggridges_0.4.1 class_7.3-14
## [7] modeltools_0.2-21 mclust_5.4 rprojroot_1.3-2
## [10] htmlTable_1.11.2 base64enc_0.1-3 proxy_0.4-21
## [13] rstudioapi_0.7 DRR_0.0.3 flexmix_2.3-14
## [16] prodlim_1.6.1 mvtnorm_1.0-7 lubridate_1.7.3
## [19] ranger_0.9.0 codetools_0.2-15 splines_3.4.4
## [22] R.methodsS3_1.7.1 mnormt_1.5-5 robustbase_0.92-8
## [25] knitr_1.20 tclust_1.3-1 RcppRoll_0.2.2
## [28] Formula_1.2-2 caret_6.0-78 ica_1.0-1
## [31] broom_0.4.3 ddalpha_1.3.1.1 cluster_2.0.6
## [34] kernlab_0.9-25 R.oo_1.21.0 sfsmisc_1.1-2
## [37] compiler_3.4.4 backports_1.1.2 assertthat_0.2.0
## [40] lazyeval_0.2.1 lars_1.2 acepack_1.4.1
## [43] htmltools_0.3.6 tools_3.4.4 igraph_1.2.1
## [46] gtable_0.2.0 glue_1.2.0 reshape2_1.4.3
## [49] Rcpp_0.12.16 trimcluster_0.1-2 gdata_2.18.0
## [52] ape_5.0 nlme_3.1-131.1 iterators_1.0.9
## [55] fpc_2.1-11 psych_1.7.8 timeDate_3043.102
## [58] gower_0.1.2 stringr_1.3.0 irlba_2.3.2
## [61] gtools_3.5.0 DEoptimR_1.0-8 MASS_7.3-49
## [64] scales_0.5.0 ipred_0.9-6 parallel_3.4.4
## [67] RColorBrewer_1.1-2 yaml_2.1.18 pbapply_1.3-4
## [70] gridExtra_2.3 segmented_0.5-3.0 rpart_4.1-13
## [73] latticeExtra_0.6-28 stringi_1.1.7 foreach_1.4.4
## [76] checkmate_1.8.5 caTools_1.17.1 lava_1.6
## [79] dtw_1.18-1 SDMTools_1.1-221 rlang_0.2.0
## [82] pkgconfig_2.0.1 prabclus_2.2-6 bitops_1.0-6
## [85] evaluate_0.10.1 lattice_0.20-35 ROCR_1.0-7
## [88] purrr_0.2.4 bindr_0.1.1 labeling_0.3
## [91] recipes_0.1.2 htmlwidgets_1.0 tidyselect_0.2.4
## [94] CVST_0.2-1 plyr_1.8.4 magrittr_1.5
## [97] R6_2.2.2 gplots_3.0.1 Hmisc_4.1-1
## [100] dimRed_0.1.0 sn_1.5-1 withr_2.1.2
## [103] pillar_1.2.1 foreign_0.8-69 mixtools_1.1.0
## [106] survival_2.41-3 scatterplot3d_0.3-41 nnet_7.3-12
## [109] tsne_0.1-3 tibble_1.4.2 KernSmooth_2.23-15
## [112] rmarkdown_1.9 grid_3.4.4 data.table_1.10.4-3
## [115] FNN_1.1 ModelMetrics_1.1.0 metap_0.8
## [118] digest_0.6.15 diptest_0.75-7 numDeriv_2016.8-1
## [121] tidyr_0.8.0 R.utils_2.6.0 stats4_3.4.4
## [124] munsell_0.4.3