Table 1 IMPROVE-IT PGx GWAS data analysis results: R2, p-values of two-sided test, and effect sizes

From: Pharmacogenomics polygenic risk score for drug response prediction using PRS-PGx methods

 

Two armsa

T armb

C armc

Two arms

T arm

C arm

Two arms

T arm

C arm

PRS method

R2

R2

R2

PvalG×T

PvalG

PvalG

\({\widehat{\beta }}_{G\times T}\) (SE)

\({\widehat{\beta }}_{G}\) (SE)

\({\widehat{\beta }}_{G}\) (SE)

PRS-Dis-CT

0.165

0.152

0.191

0.041

3.0e−09

5.1e−17

−0.031 (0.015)

−0.066 (0.011)

−0.035 (0.004)

PRS-Dis-LDpred2

0.174

0.165

0.201

0.033

4.3e−13

6.1e−23

−0.037 (0.017)

−0.079 (0.011)

−0.042 (0.004)

PRS-PGx-Unadj

0.165

0.180

0.121

0.028

1.2e−13

2.7e−03

−0.061 (0.028)

−0.082 (0.011)

0.057 (0.019)

PRS-PGx-CT

0.184

0.241

0.070

0.009

1.7e−15

0.01

−0.095 (0.036)

−0.104 (0.013)

−0.040 (0.016)

PRS-PGx-GL

0.181

0.203

0.123

0.014

7.4e−15

1.8e−03

−0.076 (0.031)

−0.093 (0.012)

0.112 (0.036)

PRS-PGx-Bayes

0.214

0.277

0.194

5.4e−05

3.8e−21

1.0e−17

−0.131 (0.032)

−0.124 (0.013)

0.198 (0.023)

  1. aTwo-arm model: Y ~ T + Sprog + T × Spred.
  2. bT-arm model: Y ~ SPGx[ = Y ~ (Sprog + Spred)], where SPGx = Sprog + T × Spred.
  3. cC-arm model: Y ~ SPGx[ = Y ~ Sprog], where SPGx = Sprog + T × Spred.