get_active_signatures: Most active signature retrieval

In greenelab/ADAGEpath: ADAGE-based Signature Analysis

Most active signature retrieval

Description

Returns signatures that are most differentially active in the limma result. We provide three methods to define "most" active: sort by the difference in activities (method = "diff"), filter by a significance level (method = "pvalue"), and consider both difference and significance using pareto fronts (method = "pareto"). This function should be used after build_limma() function.

Usage

get_active_signatures(limma_result, method = "pareto", pheno_group = "both",
  N_fronts = 5, N_signatures = 10, significance_cutoff = 0.05)

Arguments

limma_result	a data.frame that stores the limma result table returned by the build_limma() function. It should have rownames that specifies the name of each feature.
method	character, can be "diff", "pvalue", or "pareto" (default: "pareto")
pheno_group	character, can be "both", "level1", or "level2". If "both", signatures active in both phenotypes will be merged, sorted, and considered together. If "level1", only signatures active in the first phenotypes level (signatures with negative logFC) will be returned. Similarly, "level2" will return only signatures active in the second phenotypes level (signatures with positive logFC). (default: "both")
N_fronts	int, number of pareto fronts to return in the "pareto" method (default: 5)
N_signatures	int, number of top signatures to return in the "diff" method (default: 10)
significance_cutoff	numeric, significance cutoff to use in the "pvalue" method (default: 0.05).

Value

a vector storing active signatures.

greenelab/ADAGEpath documentation built on May 25, 2022, 7:11 a.m.