Function for removing low expressed genes

Question

[M K]

Dear All,

I have a matrix contains the gene names at the first columns and the rest of columns in this matrix represent the gene expression in different tissues. My question is how to write a function or line in R to remove the rows that have expression less than specific value say 5.

Answer 1

[5utr]

If you want to remove all the rows that never have expression of 5 or higher:

matrix[apply(matrix[,-1],1,function(X) length(X[X>=5])>0 ),]

[M K]

Thanks Gian.

[5utr]

If this answers your question please accept the answer

Answer 2

[h.mon]

edgeR approach filters by counts per million (CPM), see for example page 11 of the user guide. y is a DGEList:

keep <- rowSums(cpm(y)>1) >= 2
y <- y[keep, , keep.lib.sizes=FALSE]

Answer 3

[Devon Ryan]

Have a look at the genefilter package in Bioconductor. In particular, note the kOverA() function.

[M K]

I just need simple R line because I am using this file for testing something else (i.e easy line to do this regardless the first column if it is gene name oe anything else)