This variant, which has a 3.4% allele frequency in 1000 genomes data, is reported to be associated with autoimmune disease when homozygous. The statistical significance of the findings is unclear. Assuming it is pathogenic, we estimate that the relative risk of autoimmune disease in homozygous individuals is around 4x (8% risk vs. an average of 2% risk).
Surolia, I. et al. Nature advance online publication doi:10.1038/nature09115 (2010). — This paper implicates this polymorphism as potentially contributing to autoimmune diseases when homozygous. The variant was seen homozygously in 8 out of 923 cases and 0 out of 648 controls. Using two-tailed Fisher’s Exact, this is p = 0.024. However, because there are several other polymorphisms, multiple hypotheses were implicitly being tested and so we give this a case/control significance of 1 point (0.1 < p < 0.05). The overall allele frequency of the variant in controls is 0.049, they list four other variants with allele frequencies > 0.025.
They report that the protein product with this variant is catalytically active but is not secreted. This finding was reproduced in a pulse-chase experiment with 35S-methionine.
This data suggests homozygosity for this variant explains ~1% of autoimmune cases. An allele frequency of 0.049 implies that .24% of people are homozygous for this variant. Assuming an incidence of autoimmune disease in 1 out of 50 people, 0.02% of people would be expected to have autoimmune disease that might be attributed to this variant — this puts an estimate of penetrance for autoimmune disease in the homozygous individuals at ~8% (.02 / .24), approximately fourfold relative risk.