Table 1.
Estimated coefficients for the Polygenic Index (PGI) in the baseline scenario (no genetic nurture, no assortative mating; between-family analyses only)
| GWAS | Prediction sample | Meta-analysis | ORIV | PGI-RC | PGI-RC |
|---|---|---|---|---|---|
| (Default) | (GREML unc.) | ||||
| ~EA2 | N = 1000 | 0.209 | 0.522 | 0.414 | 0.414 |
| (0.147–0.271) | (0.015–1.029) | (0.287–0.542) | (0.000–0.828) | ||
| ~EA2 | N = 4000 | 0.205 | 0.501 | 0.497 | 0.497 |
| (0.174–0.237) | (0.294–0.708) | (0.421–0.573) | (0.339–0.655) | ||
| ~EA2 | N = 16,000 | 0.204 | 0.500 | 0.500 | 0.500 |
| (0.188–0.219) | (0.403–0.598) | (0.462–0.539) | (0.443–0.558) | ||
| ~EA4 | N = 1000 | 0.392 | 0.505 | 0.472 | 0.472 |
| (0.334–0.450) | (0.417–0.593) | (0.402–0.542) | (0.024–0.920) | ||
| ~EA4 | N = 4000 | 0.386 | 0.497 | 0.500 | 0.500 |
| (0.357–0.415) | (0.453–0.541) | (0.463–0.538) | (0.381–0.619) | ||
| ~EA4 | N = 16,000 | 0.387 | 0.499 | 0.500 | 0.500 |
| (0.373–0.402) | (0.477–0.521) | (0.481–0.519) | (0.456–0.544) |
Notes: Data are presented as the estimated coefficients +/− 1.96 times the standard error (95% confidence interval) for the Polygenic Index (PGI) using OLS regression on a meta-analysis based PGI (column 3), a 2SLS regression using ORIV (column 4), the default PGI-RC procedure (column 5), and the PGI-RC procedure, taking into account uncertainty in the GREML estimates (column 6). The GWAS discovery sample is set such that the resulting meta-analysis PGI has an R2 of 4.2% (~EA2) and 15.4% (~EA4). The true coefficient is 0.5. The simulation results are based on 100 replications.