Why siblings can produce offspring with each other

Last year I mentioned in passing accounts of the “Colt” family in Australia. Though the press has sensationalized them as an “incest cult,” it does seem that many of the individuals of this family are the products of sexual relations between first degree relatives (e.g., sibling or parent-child). I noted that of the children placed in foster care the one who is known to likely have been the product of out-breeding was the one who quickly began to flourish. In contrast many of the other children exhibited a variety of illnesses which likely had a genetic disposition. With all that said, please note that these first degree relatives could indeed produce offspring. This should give us a general sense that the deleterious recessives which presumably will be unmasked at high frequencies in these cases can’t be that numerous within the genome of a given individual, or the fertility of first degree crosses would be far lower.

Though there has been earlier work in this area a new preprint on arXiv, An estimate of the average number of recessive lethal mutations carried by humans, attempts to infer the average load per person using an incredibly extensive Hutterite pedigree. In the results they infer that there is fewer than 1 deleterious lethal per person in the human population. They admit with extensive caveat that this is likely a low bound, but it is probably in the right range (e.g., another result gives a figure greater by a factor of two). With such a low frequency it isn’t implausible that two individuals would lack deleterious lethal alleles altogether. One result which surprised me was the relatively low power of selection to purge genetic load even across many generations of inbreeding (something that you see most powerfully in selfing plant lineages). Of course just because someone lacks lethal mutations does not imply that they lack all alleles which express in recessive phenotypes and are deleterious. Additionally, heterozygote masking of the recessive phenotype is often incomplete due to partial dominance (at least in Drosophila). So inbreeding of near kin can still result in very serious pathologies, even if they don’t result in early mortality.