Table 3 Comparison of CV R2 of BRNSNG tissue for valid gene expression imputation models given by training base models with GTEx V8 reference panel of multiple tissues, validation base models, SR-TWAS models, and Avg-valid + SR models

From: SR-TWAS: leveraging multiple reference panels to improve transcriptome-wide association study power by ensemble machine learning

 

Sample size

Median CV R2

Mean CV R2

Ngenes

TIGAR_GTEx_BRNACC9a

136

0.040

0.062

23068

TIGAR_GTEx_BRNCDT9a

173

0.034

0.063

23362

TIGAR_GTEx_BRNCTXA9a

184

0.033

0.066

23379

TIGAR_GTEx_BRNNCC9a

182

0.033

0.060

23382

TIGAR_GTEx_BRNPTM9a

154

0.037

0.062

22481

TIGAR_GTEx_BLOOD9a

574

0.020

0.071

16682

TIGAR_GTEx_BRNSNG9b

101

0.050

0.068

23051

PrediXcan_GTEx_BRNSNG7,8,18b

101

0.052

0.080

2559

SR-TWAS_GTEx_BRNSNG

-

0.068

0.094

22913

Avg-valid + SR

-

-

-

21683

  1. aBase training model used by SR-TWAS;
  2. bBase validation model.