write_sig_genes_flexible: Use adaptive thresholds to export a set of significantly...

In BenaroyaResearch/limmaTools: Helper functions for limma analysis and visualization of results

Description

This function outputs lists of differentially expressed genes, according to a set of criteria. It uses the adjusted p-values and log-fold-changes, testing less stringent thresholds until a minimum number of genes are included. It is a wrapper for write_sig_genes.

Usage

write_sig_genes_flexible(
  topGenes, file_prefix,
  method=c("ranked_list", "combined", "directional"),
  adj_p_cut=0.01, fc_cut=log2(1.5), fc_adj_factor=1,
  p_col="P.Value", adj_p_col="adj.P.Val", fc_col="logFC")

Arguments

topGenes	a data frame, typically the output of a call to topTable. Must contain genes, log2 fold-change, and adjusted p-values.
file_prefix	name of the destination for files. Details of each output will be appended to this prefix.
method	character, specifying the type of gene lists to output. "ranked_list" outputs a list of all genes in ranked order by p-value (from smallest to largest). "combined" outputs a list of all significant genes meeting the threshold. "directional" outputs lists of significant genes meeting the threshold that are up- and down-regulated. Partial matches are allowed.
adj_p_cut	numeric, the cutoff for adjusted p-value. Genes with adjusted p-values greater than or equal to this value are not included in the result. Defaults to 0.01.
fc_cut	numeric, the absolute value cutoff for log2 fold change. Genes with absolute value log2-FC less than or equal to this value are not included in the result. Defaults to log2(1.5). To include all genes, set to 0.
fc_adj_factor	numeric, the adjustment factor used for log2-fold-change values with a numeric predictor. This is included so that the output file names can include the fold change prior to scaling. Defaults to 1, which is the appropriate value for categorical comparisons.
p_col	name or number of the column in topGenes on which to sort. Generally the raw p-values, as adjusted p-values are often homogenized across a range of raw p-values. Defaults to "P.Value", which corresponds to the output from topTable. To include all genes, set to >1.
adj_p_col	name or number of the column in topGenes containing the p-values to compare to p_cut. Defaults to "adj.P.Val", which corresponds to the output from topTable.
fc_col	name or number of the column in topGenes containing the fold-change values to compare to fc_cut. Defaults to "logFC", which corresponds to the output from topTable.

Details

This function writes out lists of genes to text files. By default, it outputs a list ranked by p-value, lists of genes significant based on FDR and logFC thresholds (all, up, and down).

BenaroyaResearch/limmaTools documentation built on Dec. 17, 2021, 10:49 a.m.