Table 2.
SNPs associated with pancreatic cancer at genome-wide significance (p < 1 × 10 7) before additional QC
| SNP | Chr. | Analysis using additional controls (p) | Differential missingness (p) | PanScan analysis (p) | Additional controls vs. MSKCC controls (p) |
|---|---|---|---|---|---|
| rs7503953 | 17 | 2.7 × 10−12 | 7.8 × 10−5 | 0.5273 | 8.2 × 10−5 |
| rs2236479 | 21 | 8.9 × 10−23 | 0.08729 | 0.7827 | 0.003 |
| rs1975920 | 12 | 1 × 10−10 | 0.448 | 0.5081 | 0.55 |
| rs1455311 | 4 | 1.3 × 10−32 | 1 | 0.2184 | 3.5 × 10−15 |
| rs1810636 | 20 | 3.4 × 10−17 | 1 | 0.4524 | 1.5 × 10−5 |
| rs1447826 | 3 | 1.1 × 10−16 | 1 | 0.2049 | 0.0014 |
In the third column, all additional controls (control:case ratio = 20:1) were used. Differential missingness was measured by a test for differences in the missing data frequency between the two groups (p value). The p value for the PanScan analysis is obtained from published data [11, 12]. The last column compared MSKCC controls with additional controls, correcting for population structure. Chr. = Chromosome.