In jasonserviss/sp.scRNAseqData: Data and processing functions for the CIMseq.data package

An example of how counts are typically filtered using the package functions.

library(CIMseq.data)
library(printr)

Load the counts data. This would usually be done with the read.table function but here we use a small test dataset that represents typical input of HTSeq generated counts.

data(testingCounts, package = "CIMseq.data")

Move the gene names to rownames.

testingCounts <- moveGenesToRownames(testingCounts)
testingCounts

Remove the .htseq suffix from colnames.

testingCounts <- removeHTSEQsuffix(testingCounts)
testingCounts

Label samples as singlets or multiplets.

testingCounts <- labelSingletsAndMultiplets(testingCounts, LETTERS[1:5])
testingCounts

Detect non-genes reported by HTSeq.

nonGenes <- detectNonGenes(testingCounts)
nonGenes
testingCounts <- testingCounts[!nonGenes, ]
testingCounts

Detect ERCC reads.

ercc <- detectERCCreads(testingCounts)
ercc
testingERCC <- testingCounts[ercc, ]
testingCounts <- testingCounts[!ercc, ]
testingERCC
testingCounts

Detect low quality genes.

lqg <- detectLowQualityGenes(testingCounts, 18)
lqg
testingCounts <- testingCounts[lqg, ]
testingCounts

Detect low quality cells (samples). Low quality cell detection can also be done with detectLowQualityCells.ERCCfrac which uses ERCC reads to detect cells with a high fraction ERCC.

lqc.totalCounts <- detectLowQualityCells.totalCounts(testingCounts, mincount = 25)
lqc.housekeeping <- detectLowQualityCells.housekeeping(testingCounts, geneName = "ACTB", quantileCut = 0.01)
lqc <- lqc.totalCounts & lqc.housekeeping
testingCounts <- testingCounts[, lqc]
testingCounts

Convert counts to a matrix.

testingCounts <- convertCountsToMatrix(testingCounts)
testingCounts

jasonserviss/sp.scRNAseqData documentation built on Jan. 8, 2020, 11:46 a.m.