Table 3.
Significant AA-associated variation is specific to KRT82 in comparison to seven other keratin genes including four type I keratins (KRT32, KRT35, KRT39, KRT40), two type II hair keratins (KRT84, KRT85), and one type II epithelial keratin (KRT5).
| Type | Cases | Controls | Case frequency | Control frequency | P-value |
|---|---|---|---|---|---|
| LoF | |||||
| KRT82 | 19 | 88 | 2.24% | 0.56% | 2.03E−06 |
| KRT84 | 0 | 13 | 0.00% | 0.08% | 1 |
| KRT85 | 1 | 3 | 0.12% | 0.02% | 0.1906 |
| KRT5 | 0 | 1 | 0.00% | 0.01% | 1 |
| KRT32 | 0 | 4 | 0.00% | 0.03% | 1 |
| KRT35 | 0 | 3 | 0.00% | 0.02% | 1 |
| KRT39 | 0 | 16 | 0.00% | 0.10% | 1 |
| KRT40 | 0 | 29 | 0.00% | 0.19% | 0.4014 |
| LoF + missense | |||||
| KRT82 | 47 | 376 | 5.54% | 2.40% | 9.19E−07 |
| KRT84 | 25 | 428 | 2.94% | 2.74% | 0.6668 |
| KRT85 | 8 | 159 | 0.94% | 1.02% | 1 |
| KRT5 | 3 | 80 | 0.35% | 0.51% | 0.8012 |
| KRT32 | 6 | 111 | 0.71% | 0.71% | 1 |
| KRT35 | 12 | 269 | 1.41% | 1.72% | 0.5869 |
| KRT39 | 8 | 127 | 0.94% | 0.81% | 0.693 |
| KRT40 | 17 | 267 | 2.00% | 1.71% | 0.4973 |
| LoF + missense + splicing | |||||
| KRT82 | 51 | 404 | 6.01% | 2.58% | 2.18E−07 |
| KRT84 | 25 | 457 | 2.94% | 2.92% | 0.9168 |
| KRT85 | 25 | 344 | 2.94% | 2.20% | 0.999 |
| KRT5 | 6 | 144 | 0.71% | 0.92% | 0.709 |
| KRT32 | 6 | 142 | 0.71% | 0.91% | 0.7081 |
| KRT35 | 12 | 284 | 1.41% | 1.82% | 0.5053 |
| KRT39 | 8 | 127 | 0.94% | 0.81% | 0.693 |
| KRT40 | 18 | 267 | 2.12% | 1.71% | 0.3438 |
Values denote the number of rare damaging variants in seven other keratin genes: four type I keratins (KRT32, KRT35, KRT39, KRT40), two type II hair keratins (KRT84, KRT85), and one type II epithelial keratin (KRT5). None of the other keratins reach significance in any of the 3 models. Type II keratins were chosen for their similar classification and positional proximity to KRT82. The type I hair keratins are the putative binding partners of KRT82. p-value was determined by two-tailed Fisher’s exact test (FET).