Table 2:

Candidate Variants for Causality in Candidate Genes

ID	POS	Family	LOD	Gene	FREQ1	FREQ2	TF Binding	Potential TF Protein
rs564982701	157406633	59	0.7561	ARID1B	0.0168	0.0011	0.93	ATF5 CCTCTTCCTTA
rs187390535	157172983	47	0.8131	ARID1B	0.0143	0.0077	0.87	BARHL2 TAAACG
rs150018283	157322342	47	0.8132	ARID1B	0.0142	0.0015	1.0	IRF7 ACTTTCGCTTTCG TRIM63 AGTTTCACTTT
rs62435519	161536849	44	1.2865	MAP3K4	0.0164	0.0061	0.61	N/A
rs140997681	161489813	42	0.5351	MAP3K4	0.0156	0.0015	0.24	GABPA ATGACTCAGCA
rs186871831	144973790	30	0.7701	UTRN	0.0151	0.0066	0.70	N/A
rs532363235	144709986	33	0.5484	UTRN	0.0149	0.0023	0.59	N/A
rs191491353	144670343	35	0.3277	UTRN	0.0171	0.0086	0.39	NFKB1 GGGGATTCCCT, NFKB2 GGGGAATCCCC, REL GGGTTTCC
rs966382235	144001704	30	0.7693	PHACTR2	0.0153	0.0003	0.61	N/A
rs79313503	143998579	33	0.5629	PHACTR2	0.0154	0.0030	0.61	N/A
rs553447284	144051910	35	0.3388	PHACTR2	0.0142	0.0002	0.58	N/A

Legend: The best candidate causal variants for each family in the best candidate gene along the linked haplotype. All variants are intronic. Here, the headers represent: ID = rs ID of the variant, POS = the physical position in base pairs, Family = Family that the minor allele of the variant appears in, LOD = LOD score of the variant from the family in column 3, Gene = the genic location of the variant, FREQ1 = the minor allele frequency of the variant in the founders in all families in our dataset, FREQ2 = the minor allele frequency of the variant in gnomAD NFE, TF binding = the transcription factor binding probability score of the variant as calculated by RegulomeDB, with 1.0 being the highest score and corresponding to a 100% chance of the variant being a TF binding site, Potential TF Protein = Protein that is known to bind to the motif that the SNP is found in (according to RegulomeDB). The binding motif of the protein is provided below the protein name (with the SNP allele bolded).