Table 1.
A. Population assignment | ||||||
---|---|---|---|---|---|---|
WB | NBW | SA | Afr | Others | Total | |
Training set | 237,055 | 10,130 | 5,206 | 4,246 | 28,024 | 284,661 |
Validation set | 33,865 | 1,448 | 744 | 607 | 4,003 | 40,667 |
Held-out test set | 67,730 | 2,894 | 1,487 | 1,213 | 8,007 | 81,331 |
Total | 338,650 | 14,472 | 7,437 | 6,066 | 40,034 | 406,659 |
B. The number of unrelated individuals used in polygenic score training | |||||||
---|---|---|---|---|---|---|---|
Model | WB | NBW | SA | Afr | Others | Total | |
i. WB-only | 237,055 | 0 | 0 | 0 | 0 | 237,055 | |
ii. Inclusive | 237,055 | 10,130 | 5,206 | 4,246 | 28,024 | 284,661 | |
iii. Inclusive-FixN | 189,449 | 10,130 | 5,206 | 4,246 | 28,024 | 237,055 | |
iv. iMultiPop, NoAdmixed | 217,473 | 10,130 | 5,206 | 4,246 | 0 | 237,055 | |
v. iPGS+refit in Afr (wo/ interaction) | 237,055 | 10,130 | 5,206 | 4,246 | 28,024 | 284,661 | |
vi. iPGS+refit in Afr | 237,055 | 10,130 | 5,206 | 4,246 | 28,024 | 284,661 | |
vii. PRS-CSx | 270,920 | 11,578 | 5,950 | 4,853 | 0 | 293,301 | |
viii. PRS-CSx (n = 256k) | 237,055 | 10,130 | 5,206 | 4,246 | 0 | 256,637 |
(A) The number of training, validation, and test-set individuals across population groups is shown. (B) The number of individuals used to train PGS models is shown. In the iPGS+refit in Afr models (models v and vi), ntrain = 284,661 individuals were used to train the iPGS model, whereas a subset of n = 4,853 individuals were used in the population-specific refit model (material and methods). Abbreviations are as follows: WB, White British; NBW: non-British White; SA, South Asian; and Afr, African.