findCandidates: The goal is to pre filter features using limma, with the hope...

In PNNL-CompBio/amlresistancenetworks: Calculates networks that lead to AML resistance to drugs

The goal is to pre filter features using limma, with the hope that the smaller set of features consist of features which are more likely to produce good models by glmnet.

Description

The goal is to pre filter features using limma, with the hope that the smaller set of features consist of features which are more likely to produce good models by glmnet.

Usage

findCandidates(mat, tmat, yvar, tyvar, p.val = 0.001)

Arguments

mat	matrix samples as rows, features as columns
tmat	see mat
yvar	(auc) response. This is what we eventually want to predict, but here we use it to define categorical variables and train a model.
tyvar	see yvar
p.val	cutoff when filtering

PNNL-CompBio/amlresistancenetworks documentation built on Feb. 8, 2025, 11:27 a.m.