Table 3 The predicted AUC values for C+T, LDpred, and LDpred2 with different parameters in the simulation experiment based on the UKBB data

From: Tuning parameters for polygenic risk score methods using GWAS summary statistics from training data

C+T

Threshold

1

5e − 1

5e − 2

5e − 3

5e − 4

5e − 5

5e − 6

5e − 7

5e − 8

  

PRStuning

0.789

0.790

0.816

0.830

0.834

0.835

0.835

0.835

0.835

  

Testing

0.793

0.795

0.830

0.852

0.860

0.865

0.867

0.868

0.868

  

LDpred

π

1

3e − 1

1e − 1

3e − 2

1e − 2

3e − 3

1e − 3

3e − 4

1e − 4

3e − 5

1e − 5

PRStuning

0.747

0.784

0.732

0.593

0.550

0.532

0.519

0.513

0.509

0.504

0.502

Testing

0.790

0.813

0.749

0.588

0.548

0.527

0.518

0.509

0.506

0.500

0.506

LDpred2

π

1

3e − 1

1e − 1

3e − 2

1e − 2

3e − 3

1e − 3

3e − 4

1e − 4

3e − 5

1e − 5

PRStuning

0.768

0.777

0.767

0.781

0.767

0.837

0.890

0.919

0.875

0.915

0.926

Testing

0.780

0.787

0.793

0.785

0.771

0.818

0.846

0.859

0.838

0.856

0.861

  1. We randomly selected 80% of individuals as training data and the rest as the testing data. The data from the 1KG were used as the reference panel. For C+T, LDpred, and LDpred2, respectively, the correlation values ρAUC reached 0.994, 0.998, 0.989, the relative difference values rdAUC were 3.8%, 1.3%, 7.0%, and the sensitivity values of the tuned PRS model based on PRStuning were 80.6%, 74.8%, 85.3%. Sensitivity is calculated using the cutoff value selected by Youden’s J statistic.