bic.deseq.normalize.htseq.counts: Use DESeq to normalize raw HTSeq counts

In bic-mskcc/bicrnaseq: MSKCC Bioinformatics Core RNASeq

Description

Use DESeq scaling method to normalize raw HTSeq counts and return a list containing both raw and normalized counts. Takes in either formatted.counts matrix from bic.format.htseq.counts() OR an HTSeq counts file. Optionally takes in previously generated countDataSet. If not given, also optionally takes in parameters for generating one. Sample key will be used in generation of countDataSet but if not given, it will be assumed that all samples belong to one condition group.

Usage

bic.deseq.normalize.htseq.counts(formatted.counts = NULL, htseq.file = NULL,
  cds = NULL, key = NULL, min.count = 10, libsizeQ = F,
  percentile = "100%", fitType = "parametric", method = "per-condition",
  sharingMode = "maximum")

Arguments

formatted.counts	matrix containing raw counts from HTSeq. One column per sample, one row per gene; may be used INSTEAD of htseq.file, but NOT in conjunction with
htseq.file	file containing raw HTSeq counts; One column per sample, one row per gene; may be used INSTEAD of formatted counts, but NOT in conjunction with
cds	DESeq countDataSet object; Default: NULL, if null will compute on the fly
key	sample key where column one contains sample ID that matches that in counts file, and column two contains the condition group to which it belongs. Only those samples that are in the key matrix will be included in the returned matrices
min.count	minimum total number of reads for a gene across samples; Default: 10
libsizeQ	logical to indicate we should use library size to normalize counts rather than DESeq's method [TO DO: RECHECK THIS]
percentile	string indicating percentile to use when normalizing by library size
fitType	See DESeq documentation. Default: "parameteric"
method	See DESeq documentation. Default: "per-condition"
sharingMode	See DESeq documentation. Default: "maximum"#'

Value

list containing two matrices: one of raw counts and one of DESeq-scaled, log2 transformed counts. Column names will have condition appended

bic-mskcc/bicrnaseq documentation built on May 24, 2019, 3:04 a.m.