The criteria are inadequate for predicting an A blood type.
If an individual has (rs8176746(G;G), rs8176747(C;C), and rs8176719 (D;I), he or she may in fact have two O alleles, instead of AO, provided that one of the O alleles is non-deletional. The non-deletional O allele O303 is sufficiently common to warrant consideration. O303, or ABO*O03 (O2), is one of "(t)he 6 common ABO alleles in white individuals", along with ABO*A101 (A1), ABO*A201 (A2), ABO*B101 (B1), ABO*O01 (O1), and ABO*O02 (O1v). (http://www.bloodjournal.org/content/102/8/3035?sso-checked=true)
To exclude this possibility, (rs41302905)(C;C) -- (rs41302905)(G;G) on the reverse strand -- should be added to the existing criteria.
(This is not hypothetical; I am in precisely this situation. I have (rs8176746(G;G), rs8176747(C;C), and rs8176719(D;I), which causes me to be predicted by Promethease as type A (AO). However, I also have (rs41302905)(C;T). My ABO alleles as predicted by 23andMe are O112 and O303. My blood type, as determined on multiple occasions through conventional serological typing, is O+.)
- Thanks for the info. I've added ' not(rs41302905(T)) ' to the 'and', which I think will improve this, while allowing it to still showup if someone has no data for rs41302905.
- That is however slightly different from your suggestion, in how it handles heterozygosity at rs41302905. How should a rs41302905(C;T) be called?
- If you see it needing further help, I'm interested. email firstname.lastname@example.org if you'd like some coupons for free promethease runs so that you can test it.
- --- cariaso 20:34, 7 February 2015 (UTC)
This is good, I think; thanks. As far as I have been able to determine, the majority of those who do have rs41302905(C;T) probably have O303. Another non-deletional O might be possible, and in still-less-common situations, even perhaps a weak A. But compared to the other possibilities, O303 seems to be the most common. (See the link I posted above.)
Among other things, this means that if an individual has both rs8176719(D;I) and rs41302905(C;T), there is a probability that the individual will be type O (OO). One allele will be a more typical deletional O allele; the other will be a non-deletional O. Of course, it seems that will ABO alleles and blood type, there are always surprises.