From: Rare variants analysis using penalization methods for whole genome sequence data
Model-1: | 10 % of variants in influential regions are causal with effect size +1, while each one is correlated with some neutral variants in their region. |
Model-2: | 10 % of variants in influential regions are causal with effect size ±1, while each one is correlated with some neutral variants in their region. |
Model-3: | 10 % of variants in influential regions are causal with effect size +1, while they are uncorrelated with other variants in their region. |
Model-4: | 10 % of variants in influential regions are causal with effect size ±1, while they are uncorrelated with other variants in their region. |
Model-5: | 20 % of variants in influential regions are causal with effect size selected from U(0.5,1), while 10 % are correlated and 10 % are uncorrelated with other causal and neutral variants in their region. |
Model-6: | 20 % of variants in influential regions are causal with effect size selected from U(−1,−0.5) and U(0.5,1) while 10 % are correlated and 10 % are uncorrelated with other causal and neutral variants in their region. |