pagoda.subtract.aspect: Control for a particular aspect of expression heterogeneity...

In hms-dbmi/scde: Single Cell Differential Expression

Control for a particular aspect of expression heterogeneity in a given population

Description

Similar to subtracting n-th principal component, the current procedure determines (weighted) projection of the expression matrix onto a specified aspect (some pattern across cells, for instance sequencing depth, or PC corresponding to an undesired process such as ribosomal pathway variation) and subtracts it from the data so that it is controlled for in the subsequent weighted PCA analysis.

Usage

pagoda.subtract.aspect(varinfo, aspect, center = TRUE)

Arguments

varinfo	normalized variance info (from pagoda.varnorm())
aspect	a vector giving a cell-to-cell variation pattern that should be controlled for (length should be corresponding to ncol(varinfo$mat))
center	whether the matrix should be re-centered following pattern subtraction

Value

a modified varinfo object with adjusted expression matrix (varinfo$mat)

Examples

data(pollen)
cd <- clean.counts(pollen)

knn <- knn.error.models(cd, k=ncol(cd)/4, n.cores=10, min.count.threshold=2, min.nonfailed=5, max.model.plots=10)
varinfo <- pagoda.varnorm(knn, counts = cd, trim = 3/ncol(cd), max.adj.var = 5, n.cores = 1, plot = FALSE)
# create go environment
library(org.Hs.eg.db)
# translate gene names to ids
ids <- unlist(lapply(mget(rownames(cd), org.Hs.egALIAS2EG, ifnotfound = NA), function(x) x[1]))
rids <- names(ids); names(rids) <- ids
go.env <- lapply(mget(ls(org.Hs.egGO2ALLEGS), org.Hs.egGO2ALLEGS), function(x) as.character(na.omit(rids[x])))
# clean GOs
go.env <- clean.gos(go.env)
# convert to an environment
go.env <- list2env(go.env)
# subtract the pattern
cc.pattern <- pagoda.show.pathways(ls(go.env)[1:2], varinfo, go.env, show.cell.dendrogram = TRUE, showRowLabels = TRUE)  # Look at pattern from 2 GO annotations
varinfo.cc <- pagoda.subtract.aspect(varinfo, cc.pattern)

hms-dbmi/scde documentation built on April 19, 2023, 10:21 p.m.