Rank Genes by Correlation With a Deconvolution Subgroup

Computes a data-driven gene ranking by correlating expression values with a selected deconvolution subgroup across matched samples.

Usage

compute_data_driven_rank(res, expr, deconv, subgroup, method = "spearman")

Arguments

res

Data frame containing at least a gene column with genes to evaluate.

expr

Gene expression matrix with genes in rows and samples in columns.

deconv

Deconvolution matrix with samples in rows and features/subgroups in columns.

subgroup

Character scalar indicating the subgroup/feature column in deconv.

method

Correlation method passed to stats::cor(), typically "spearman" or "pearson".

Value

A data frame sorted in descending correlation, with one column named correlation and row names corresponding to genes.