. 2019 Apr 15;7:e6755. doi: 10.7717/peerj.6755

Table 3. Results of single SNP analysis and SNP × gender interaction analysis.

SNP	Total insight problem solving				Verbal insight problem solving				Figural insight problem solving
	Genotype		Genotype × Gender		Genotype		Genotype × Gender		Genotype		Genotype × Gender
	Test statistic	p_emp	Test statistic	p_emp	Test statistic	p_emp	Test statistic	p_emp	Test statistic	p_emp	Test statistic	p_emp
rs737865	0.421	.811	0.382	.826	2.78	.251	1.24	.543	1.50	.470	0.234	.889
rs5993883	0.793	.667	0.541	.766	0.938	.632	0.653	.729	0.767	.685	1.94	.374
rs6269	0.639	.726	0.343	.840	0.073	.964	0.826	.658	2.19	.334	0.015	.992
rs4633	0.447	.805	3.90	.146	1.60	.449	4.46	.113	0.797	.672	1.59	.460
rs6267	−0.359	.733	1.83	.175	−0.316	.748	0.350	.548	−0.357	.720	2.29	.133
rs4818	0.634	.726	0.241	.886	0.100	.953	0.408	.806	1.89	.391	0.049	.975
rs4680	0.204	.905	2.31	.315	1.84	.409	2.72	.255	0.518	.774	1.13	.578

Notes.

Single SNP effect and SNP × gender interaction were tested under the genotypic model using linear regression with a 2df joint test. For SNP (rs6267) with minor allele homozygotes <5%, minor allele homozygote carriers and heterozygote carriers were collapsed into one group (the dominant model) for analysis. Empirical p values (p_emp) were obtained by 10,000 permutations.