I commented on the "throwback family" a few days ago, well, The Times (of London) has twoarticles which reduce the likelihood of this being a hoax in my mind. It seems clear that there is a family, highly inbred, which lives in Turkey where a number of the children walk on all fours and exhibit other forms of impairment. Nevertheless, the exact details of what is going on here is problematic to me. As usual, the newspapers tend to garble and confuse many issues. Consider this sentence: "All five are brain-damaged because of a mutation in a gene 17p, located on chromosome 17, which occurred because mother and father are closely related: Hatice's maternal grandmother was the sister of Resit's father (which makes them one-sixteenth related)." This is likely poor grammar, but the mutation did not occur because of a close genetic relationship. That is, a high degree of inbreeding does not spontaneously increase the rate of mutation, rather, consanguineous marriages tend to increase the chances that deleterious mutations will be unmasked. The principle is simple. Consider an assertion sometimes made that every human being has 3 lethal recessives. The point here is that if the allele, the genetic variant, were carried as two copies (humans are diploid, we have two copies of all genes) so that the individual was homozygous, than negative phenotypic consequences would ensue and reduce fitness. So, cystic fibrosis is a lethal recessive, if a long acting one (with medication today CF patients can survive deep into their 30s). 1 out of 20 Western Europeans carry the CF mutation. Of these individuals, almost all of them are heterozygous, that is, they carry one "good copy" and a "bad copy." These individuals do not manifest any negative phenotypic outcomes because the good copy of the gene is sufficient for normal functioning. The problem occurs when two individuals who carry a copy of the gene mate, there is a 25% chance that the child will receive both bad copies and so the negative phenotypic consequences will be unmasked. If 1 out of 20 individuals in the population is heterozygous, you have a random mating chance of 1/400 of carriers pairing up, and a 1/4 chance that any resultant offspring will exhibit CF. So, consider a situation where you have one parent who is a CF carrier. This individual's children with a homozygous "wild type" individual (no bad copies) will have a 1/2 chance of being a barrier. Now, imagine that the children mate with each other. Each of the children has a 1/2 chance of being a CF carrier, so you have a 1/4 chance two carriers will mate, and a 1/4 chance that offspring with negative phenotypic consequences will result. Multiplying out the independent probabilities implies that if a carrier's children mate incestuously the expectation of a grandchild exhibiting CF is 1/8. Compare this to someone whose children do not mate incestuously, each of the children have a 1/2 chance of being carriers, and each of these have a 1/20 chance of mating another carrier, so you have a 1/40 chance of having a double carrier pairing. 1/4 of the children will exhibit the phenotype, so you have a 1/160 chance of a grandchild of this individual having CF. In other words, the probability difference between sibling matings and non-incestuous matings of resulting in a CF grandchild for a heterozygous individual is a factor of 20. This is an extreme situation, but multiple generations of mild incest can be problematic, in other words, I am skeptical that the coefficient of relatedness is simply 1/16. Again, unfortunately, the atavism angle is emphasized. This is good copy, but we should be skeptical of this. It is possible, but that does not imply it is probable. It seems that the 17p mutation is a loss of function which has had wide pleiotropic effects. I am skeptical that the ancestral gene shifted from null function to a new function. As Carl Zimmer pointed out bipedalism is a complex and exceptional trait which resulted in the retrofitting of a host of other aspects of our morphology and physiology. Though it might not have emerged in a classic gradualistic fashion, it seems unlikely to have been the result of a single positive mutation around which modifier genes evolved over hundreds of generations. So, if the gene that is causing this non-bipedalism is of a different nature on the sequence level from the character of our ancient pre-hominin ancestors, what exactly can be infer about the evolution of bipedalism from this single family? I don't know. Finally, there is the issue that one of the children in this family who carries the 17p mutation does not seem to exhibit the aberrant traits of his siblings. The lead scientist who is quoted in the article suggests that this is both nature and nurture, that a particular environmental context increased the probability of this phenotype expressing itself given the initial genotype. In the technical literature this would probably be termed incomplete penetrance. But, I want to zoom in on the conclusion and question how deeply Humphrey has thought this through, or, how much we should trust The Times if it has taken him out of context:
Professor Humphrey says that a combination of factors - the genetic defect, the huge family, limited parental attention, sibling imitation, skeletal anomalies, the lack of intervention from the outside world - led to this remarkable outcome. "If there was just a one in a thousand chance of each of these things happening, this soon adds up to a one in a billion chance of it happening in the same family. It's not surprising we've never seen it before. We may never see it again."
In the first paragraph there are a host of variables listed, and in the second Humphrey seems to be implying that these variables are independent, that each offspring is an independent trial which can be viewed separately. I don't see how this works, it seems that many of the variables listed above should exhibit some correlational and interactional effects with each other. I have shown above how the probability of a the genetic defect manifesting itself increases greatly within the same family. This family should also likely foster a similar environmental input to its offspring, even if it varies. I am probably confused, because I don't see exactly what Humphrey is trying to get at in the second paragraph in regards to probabilities. Finally, the nature and nurture in conjunction might be correct, but, there is an alternative explanation for incomplete penetrance, and that is that traits are often polygenic in a way we do not comprehend because of an epistatic, gene-to-gene interaction, parameter within the equation. In other words, the "normal" son might carry a gene on another locus which masks the non-bipedal phenotype. In short, I'm convinced this isn't a hoax, and it might be medically interesting. But I am also skeptical of the claim that this tells us something on a broad evolutionary genetic and palaeoanthropological scale.