okay no actually what is the hardy weinburg expected ratio of labrador colors if we pretend purely random mating and no humans making color choices? i.e., how common should chocolate labs be under Hardy Weinburg?
I found a study (open access, even!) that lists the allele frequencies of Labs in three separate populations (UK show, n=104; US field, n=175; US show, n=92) and also pools them in one pan-Lab sample (n=371) for easy analysis. awesome. (okay so I have to scrape those numbers out of the supplementals, but shh.) that study puts the ratio of the e allele at 0.60 overall: 0.46 for UK show, 0.61 for US field, and 0.75 (!) for US show). excellent.
okay, and apparently there are three synonymous b alleles that can create liver in labs at TYRP1, but they all act exactly the same. so. uh, that's something. for simplicity, let's calculate the allele frequency for B instead...
obviously not all dogs genotyped at all locations, so the ns and allele frequencies for the TYRP1 locus at B for black are... UK show 0.64 (n = 107), US field 0.49 (n = 177), US show 0.83 (!) (n = 92)... with a total sample of 376 dogs here and a combined allele frequency of 0.61 B alleles. we can translate that to say all the non-black alleles (so the remaining fraction needed to make 100% of all the alleles) shakes out with these numbers for the pan-b allele we really care about:
UK show: 0.36
US field: 0.51
US show: 0.17
overall: 0.39
since the formula for the overall proportion of chocolate dogs would be p(bb) * p(B_) (probability of two copies of b = liver AND probability of at least one copy of E for not-yellow)...
easier to score this as p(b allele)^2 (which is the HW phenotypic relationship) and 1-p(y)^2 for the probability of NOT being yellow...
so that would come out with ratios of chocolate phenotypes such within each subpopulation:
UK show: (0.36^2) * (1-0.46^2) = 10.2% of all labs chocolate
US field: (0.51^2) * (1-0.61^2) = 16.3 % of all labs chocolate
US show: (0.17^2) * (1-0.75^2) = 1.3% of all labs chocolate
overall: (0.39^2) * (1-0.60^2) = 9.7 % of all labs chocolate
now I suspect very much that this is NOT the case because of chocolate lab breeders and of course you can selectively breed for anything forever. but this is probably why in many circles chocolate dogs have a reputation for not being "good" dogs: purpose bred lab communities tend not to have them pop up too much and . and especially this is probably driven by how much you give a shit about "yellow chocolates" bc the US show circuit is apparently AGGRESSIVELY trying to minimize the odds there, given that the US show circuit seems to be going in very heavily on yellows.
fuck this paper is awesome actually, it has some really interesting conclusions about breed ancestry and rare allelic variants that indicate deeper relationships. fuckin love that.
