Table 1.

Summary of parameter settings in the simulation study

Setting	$K$	$M_k$	$p_k$(%)	$q_k$(%)	Overlap pattern
1	4	49,750, 37,500, 25,125, 12,625	0.5, 1, 1.5, 2	0, 0, 10, 90	Non-overlap
2	4	49,750, 37,500, 25,125, 12,625	0.5, 1, 1.5, 2	0, 0, 50, 50	Non-overlap
3	4	49,750, 37,500, 25,125, 12,625	0.5, 1, 1.5, 2	10, 20, 30, 40	Non-overlap
4	4	49,750, 37,500, 25,125, 12,625	0.5, 1, 1.5, 2	25,25,25,25	Non-overlap
5	10	12,375 × 6, 12,625, 12,625, 12,750, 12,750	0 × 6, 2, 2, 3, 3	0 × 8,10, 90	Non-overlap
6	4	49,750, 37,500, 25,125, 12,625	0.5, 1, 1.5, 2	0, 0, 10, 90	Overlap I
7	4	49,750, 37,500, 25,125, 12,625	0.5, 1, 1.5, 2	0, 0, 10, 90	Overlap II

A total of $M=125,000$ SNPs were sampled from the UK Biobank data with quality control filters, with 1 KG as LD reference panel. The sampled variants were then assigned to $K$ different annotation groups, which explain $q_1,...,q_K$% of the total heritability $h^2$ respectively. For each annotation group $k$, the proportion of causal variants is denoted as $p_k$. Overlapping pattern I was used in setting 6, where the intersection over union metric (IOU) was higher among the annotation groups with low heritability contribution. For setting 7, overlapping pattern II was used, where IOU was higher among the annotation groups with high heritability contribution