Single-marker analysis of SNPs from 4 genes on chromosome 22 by PLINK.
The major allele in the control group is designated as allele 1, and the minor allele as allele 2 unless follows allele designation in initial study.
†Four genes that showed suggestive significance in set-based test in the discovery sample set (see Table 4) were followed up with the replication sample set. Six SNPs within these four genes with nominal P<0.05 (marked by ¶) were first chosen for follow-up. Nineteen more SNPs were also selected for follow-up for the following reasons: (i) in LD with these six SNPs (not marked by any symbol); (ii) with potential functional relevance – T for SNP located at predicted transcription factor binding site, and M for SNP located at predicted microRNA binding site); or (iii) forming significantly associated haplotype windows with one of the six chosen SNPs (marked by #) – rs2238754 (#) and rs2800960 (¶) with normal P = 5.19e-13; rs9610583 (#) and rs4616572 (¶) with nominal P = 9.32e-46; and rs1217125 (#), rs2284021 (#) and rs4820254 (¶) with nominal P = 1.06e-07. As such, 13 SNPs that had not been genotyped in the discovery sample set are shown as missing data (−).
‡SNPs are listed according to their sequential physical positions on chromosome 22 (NCBI build 37.1).
§Empirical P values (Pemp) are estimated based on 10,000,000 permutations. In each round of permutation (swapping of the case-control status), the best original result of every SNP is compared against the best result of the three tests (allelic, dominant and recessive) of that SNP, and also against the best results from all SNPs.
+SNPs tested to have significant difference (P<0.05) in minor allele frequency (MAF) for control subjects between the discovery and the replication sample sets.
