Fig. 4: Accuracy of 6 prediction methods in height meta-analysis of UK Biobank and 23andMe cohorts. | Nature Communications

Fig. 4: Accuracy of 6 prediction methods in height meta-analysis of UK Biobank and 23andMe cohorts.

From: Incorporating functional priors improves polygenic prediction accuracy in UK Biobank and 23andMe data sets

Fig. 4

We report results for P + T, LDpred, P + T-funct-LASSO, AnnoPred, LDpred-funct-inf, and LDpred-funct, for each of 4 training data sets: UK Biobank interim release (113,660 training samples), UK Biobank (408,092 training samples), 23andMe (698,430 training samples) and meta-analysis of UK Biobank and 23andMe (1,107,430 training samples). Nested training data sets are connected by solid lines (e.g. UK Biobank (408 k) and 23andMe are both connected to Meta-Analysis, but not to each other). The dashed line denotes the estimate of SNP-heritability in the UK Biobank. Numerical results are reported in Supplementary Table 27. Block-Jackknife standard errors over 200 genomic jackknife blocks were < 0.028 across each method. Source data are provided as a Source Data file.

Back to article page