Skip to main content

Table 3 Simulation results with MA data using GCTA model with the true heritability \(h^2 = 0.5\)

From: Boosting heritability: estimating the genetic component of phenotypic variation with multiple sample splitting

 

100 causal SNPs

5000 causal SNPs

100 causal SNPs from 3 genes

GCTA

0.3176 (.0268)

0.2890 (.0358)

0.3614 (.0395)

Enet

0.4014 (.0393)

0.4018 (.0538)

0.3965 (.0474)

HERRA

0.5248 (.0342)

0.5142 (.0586)

0.5246 (.0427)

B_herra

0.5217 (.0260)

0.5150 (.0447)

0.5192 (.0339)

Enet (remove 60% covariates)

0.4541 (.0364)

0.4614 (.0371)

0.4408 (.0403)

HERRA (remove 60% covariates)

0.4988 (.0356)

0.4941 (.0469)

0.4966 (.0338)

B_herra (remove 60% covariates)

0.5015 (.0260)

0.4892 (.0426)

0.4965 (.0306)

  1. The mean and the standard deviation (in parentheses) of the estimated heritabilities between the simulation replicates are presented