Have questions? Visit https://www.reddit.com/r/SNPedia


From SNPedia

I have Gs220 but as a stem cell donor I know that while I am HLA-B15, I am _not_ HLA*B1502 but HLA-B*15CYFU, i.e. one of:

1501 (this is most likely, it is most widespread) 1501N 1533 1534 1560 1570 1571 1578 1579N 1581 1582 1592 1596 1597 9502 9504 9513 9522

I suspect the key is to link this with rs3909184:G, which I do not have.

[PMID 16998491] I interpret alleles 1,2,3,4 to be = to A,C,G,T based on refs like this

bottom line:

HLA-B*1502 comes about when (in phase) you have rs3909184(C) + rs2844682(T), which happens primarily in certain East Indian populations

so PharmGKB is correct, in that the common (C) allele of rs3909184 is indeed the one along with (T) for rs2844682 in the risk haplotype

clearly it's only useful to know rs3909184 in the genoset and not on it's own

meanwhile ... it's not actually clear from the cited paper whether this works only in CHB (Chinese) populations or if it works across populations (ie in CEU, etc), but the frequency of rs2844682(T) seems so low in these (other) populations that we probably don't need to worry about it

the bigger worry is the usual, in unphased data, we can't tell this person:

dad's ch: rs3909184(C) + rs2844682(T) ... and therefore carrying a B*1502
mom's ch: rs3909184(G) + rs2844682(C)

apart from this person:

dad's ch: rs3909184(C) + rs2844682(C)
mom's ch: rs3909184(G) + rs2844682(T)

and therefore not carrying a B*1502