4.13 实战十三 | 10X | 小鼠乳腺上皮细胞

刘小泽写于2020.7.21

1 前言

数据来自Bach et al. (2017)，使用的是妊娠期小鼠乳腺上皮细胞 + 10X技术建库

数据准备

library(scRNAseq)
sce.mam <- BachMammaryData(samples="G_1")
sce.mam
# class: SingleCellExperiment 
# dim: 27998 2915 
# metadata(0):
#   assays(1): counts
# rownames: NULL
# rowData names(2): Ensembl Symbol
# colnames: NULL
# colData names(3): Barcode Sample Condition
# reducedDimNames(0):
#   altExpNames(0):

数据初探

# 样本信息
sapply(names(colData(sce.mam)), function(x) head(colData(sce.mam)[,x]))
# Barcode              Sample Condition  
# [1,] "AAACCTGAGGATGCGT-1" "G_1"  "Gestation"
# [2,] "AAACCTGGTAGTAGTA-1" "G_1"  "Gestation"
# [3,] "AAACCTGTCAGCATGT-1" "G_1"  "Gestation"
# [4,] "AAACCTGTCGTCCGTT-1" "G_1"  "Gestation"
# [5,] "AAACGGGCACGAAATA-1" "G_1"  "Gestation"
# [6,] "AAACGGGCAGACGCTC-1" "G_1"  "Gestation"

ID转换

依然是整合行名 + 添加染色体信息

library(scater)
rownames(sce.mam) <- uniquifyFeatureNames(
    rowData(sce.mam)$Ensembl, rowData(sce.mam)$Symbol)

library(AnnotationHub)
ens.mm.v97 <- AnnotationHub()[["AH73905"]]
rowData(sce.mam)$SEQNAME <- mapIds(ens.mm.v97, keys=rowData(sce.mam)$Ensembl,
    keytype="GENEID", column="SEQNAME")

# 总共有13个线粒体基因
sum(grepl("MT",rowData(sce.mam)$SEQNAME))
# [1] 13

2 质控

依然是备份一下，把unfiltered数据主要用在质控的探索上

unfiltered <- sce.mam

使用线粒体信息进行过滤

is.mito <- rowData(sce.mam)$SEQNAME == "MT"
stats <- perCellQCMetrics(sce.mam, subsets=list(Mito=which(is.mito)))
qc <- quickPerCellQC(stats, percent_subsets="subsets_Mito_percent")

colSums(as.matrix(qc))
##              low_lib_size            low_n_features high_subsets_Mito_percent 
##                         0                         0                       143 
##                   discard 
##                       143

作图

colData(unfiltered) <- cbind(colData(unfiltered), stats)
unfiltered$discard <- qc$discard

gridExtra::grid.arrange(
    plotColData(unfiltered, y="sum", colour_by="discard") + 
        scale_y_log10() + ggtitle("Total count"),
    plotColData(unfiltered, y="detected", colour_by="discard") + 
        scale_y_log10() + ggtitle("Detected features"),
    plotColData(unfiltered, y="subsets_Mito_percent", 
        colour_by="discard") + ggtitle("Mito percent"),
    ncol=3
)

再看看线粒体含量与文库大小的关系

plotColData(unfiltered, x="sum", y="subsets_Mito_percent", 
    colour_by="discard") + scale_x_log10()

最后过滤

dim(unfiltered);dim(sce.mam)
# [1] 27998  2915
# [1] 27998  2772

3 归一化

使用去卷积的方法

library(scran)
set.seed(101000110)
clusters <- quickCluster(sce.mam)
sce.mam <- computeSumFactors(sce.mam, clusters=clusters)
sce.mam <- logNormCounts(sce.mam)

4 找高变异基因

这里由于是10X的数据，所以会有UMI信息，因此可以用基于泊松分布的模型构建方法

set.seed(00010101)
dec.mam <- modelGeneVarByPoisson(sce.mam)
top.mam <- getTopHVGs(dec.mam, prop=0.1)

最后做个图

plot(dec.mam$mean, dec.mam$total, pch=16, cex=0.5,
    xlab="Mean of log-expression", ylab="Variance of log-expression")
curfit <- metadata(dec.mam)
curve(curfit$trend(x), col='dodgerblue', add=TRUE, lwd=2)

5 降维聚类

降维

library(BiocSingular)
set.seed(101010011)
sce.mam <- denoisePCA(sce.mam, technical=dec.mam, subset.row=top.mam)
sce.mam <- runTSNE(sce.mam, dimred="PCA")

# 检查PC的数量
ncol(reducedDim(sce.mam, "PCA"))
## [1] 15

聚类

有一个很重要的参数是k ，含义是：the number of nearest neighbors used to construct the graph。如果k设置越大，得到的图之间联通程度越高，cluster也越大。因此这个参数也是可以不断尝试的

我们这里由于细胞数量比较多，所以设置的k就比较大，得到的cluster就少而大

snn.gr <- buildSNNGraph(sce.mam, use.dimred="PCA", k=25)
colLabels(sce.mam) <- factor(igraph::cluster_walktrap(snn.gr)$membership)

table(colLabels(sce.mam))
## 
##   1   2   3   4   5   6   7   8   9  10 
## 550 799 716 452  24  84  52  39  32  24

作图

plotTSNE(sce.mam, colour_by="label")