Table 1 Guilt-by-association methods.

From: Improving gene function predictions using independent transcriptional components

 

DEPICT and GADO

GENETICA (ours)

Strategy

Guilt-by-association prediction of gene functions using gene sets

Guilt-by-association prediction of gene functions using gene sets

Transcriptional components

Principal component analysis

Independent component analysis

Barcode

Comparison of eigenvector scalars of member genes and nonmember genes using Welch T test

Average of all mixing matrix weights of member genes

Distance metric

Pearson correlation

Distance correlation

Permutation test

No, parametric p values used

Yes, permutation p values used

Reference

5,13

This study