Monday, June 19, 2006

10 questions for David Haig   posted by Razib @ 6/19/2006 01:20:00 PM

David Haig is the editor of Genomic Imprinting and Kinship. You may find many of his publications on his website. Below are 10 questions....

1) Reviewing your work on genomic imprinting I detect some frustration with those who suggest that monandrous mating systems imply symmetrical expression of "madumnal" and "padumnal" (for readers, inherited from mother or father, respectively) copies of a particular gene. Your basic argument seems to be is that in pure monandry madumnal and padumnal copies should agree on overall expression levels, but how they get there is irrelevant (i.e., the proportion of a gene product contributed by each). This seems trivially clear once pointed out, so my question is: did the researchers to whom you were responding really read your original papers? Or were you simply
unclear on this issue?

It is my experience that the majority of working biologists do not read theoretical papers closely and, when they do, often do not put in the effort to really understand the arguments. I can understand this, given that this is not their specialty and given the pressures on all our time. Things are not helped by 'silly' statements by a few theoretical biologists who do not have this excuse.

2) R.A. Fisher and Sewall Wright disagreed on the important of statistical epistatic interactions (i.e., those that result in variation across a population) in evolutionary process. Fisher seemed to assume they were as irrelevant as population substructure and random genetic drift, while gene-gene interactions played an important role in the adaptive landscapes in Wright's Shifting Balance Theory. 1) Is this argument relevant today? (in the post Neutral Theory and postgenomic world) 2) Do you have an opinion on the importance of epistasis in evolutionary processes?

Discussion about the role of 'epistasis' is complicated because the term has different meanings in biochemistry/molecular biology and in population genetics. Biochemical epistasis occurs when two genes have products that interact in a single pathway. Population-genetic epistasis occurs when two loci are polymorphic and there is a statistical interaction between alleles at the two loci in their effects on fitness. One can have one sort of epistasis without the other, and it is not always clear which definition people are using.

There is no question that biochemical epistasis is important in evolution. There is still controversy about population-genetic epistasis. This is not my area of expertise but I suspect that the answer may change depending on an individual's implicit time-scale. I have had a particular interest in fitness interactions between genes in mothers and genes in their offspring. Such interactions have the property that mothers treat offspring differently depending on which genes they inherit. If the loci in mother and offspring are tightly linked, then such interactions can have the properties of genetic 'self-recognition' or green-beard effects.

3) Do you believe most biologists, even evolutionary biologists, appreciate formal theory?

Most biologists do not appreciate formal theory. Theory is more respected by evolutionary biologists as a group.

4) You note that there is some evidence that related species of mice with alternative reproductive strategies, polyandrous vs. monandrous, exhibit different levels of genomic imprinting confirmed by hybridizations. What is the general time scale of how fast genomic imprinting can evolve? My understanding is that sexual dimorphism emerges rather slowly because of its peculiar sex contextual nature, would genomic imprinting be similar?

I do not know how rapidly such divergences in imprinting can evolve. We really need more data on more cases of closely related species with different mating systems.

5) You make it clear that one would expect monandrous species to exhibit a wide range of madumnal and padumnal expression on imprinted loci. But, it seems to me that in regards to a monandrous species that had "relaxed constraint" upon imprinting one could ascertain an expectation of the distribution of expression ratios between the two copies across loci. In short, could this be a way to infer aspects of sexual natural history of a monandrous species? (e.g., the length and extent of monandry)

I think there may be information of the kind you mention in the pattern of imprinted gene expression. For example, it is possible that maternally-expressed and paternally-expressed genes will react differntly to changes in mating system and that this may leave an evolutionary trace.

6) In your review of Narrow Roads of Gene Land: Volume 2, you addressed the late William D. Hamilton's attempt to apply evolutionary biology to humanity and our present state. In particular Hamilton was an unabashed eugenicist. My general impression is that though you did not question Hamilton's heart, you were skeptical of the particular plans he forwarded to implement his solution to the problem of deterioration of the human genome through the lack of selection against deleterious alleles. Now, seeing as how there are about 6 1/2 billion humans on this planet, is there any plausability that 'mutational meltdown' could be an issue for our species? And in regards to our health, is not likely that since our effective population is so large there should still be a wide range of realized variance in mutational load so that a small percentage may carry the torch of "genetic health" into the future even in a scenario where our technological civilization can no longer prop up the health of the genetically suboptimal majority?

These are very complicated questions about which I am wary of giving an ill-considered answer. Hamilton was a great proponent of the importance of sexual selection in keeping the gene-pool free of deleterious mutations (survival makes no difference if you do not reproduce) but he does not appear to have given this much thought with respect to humans. Selection on reproduction may have been less relaxed than selection on survival. Personally, I think that we should be worrying foremost about the environmental problems of having 6.5 billion people (these problems are here now and will get worse soon) rather than worry about the very long-term danger of a mutational meltdown.

7) In a follow up in regards to accumulation of deleterious alleles, in The Cooperative Gene Mark Ridley suggests that spontaneous abortions are purging the genetic load from our population. If survivability and reproductive value of individuals who carry a high load of deleterious alleles in elevated in our modern population, that seems to imply to me that spontaneous abortions would be elevated as well, perhaps serving as a check on the accumulation of load. Is this possible?

It is possible.

8) In Genes in Conflict Austin Burt and Robert Trivers point out that a large number of the loci in mice which are imprinted have behavioral and neurological implications. This seems born out by Prader-Willi Syndrome in humans. Do you believe that many cognitive traits or biases will show the effects of imprinting?

I must await the evidence on this but I would not be surprised if this were true. Theory does predict that genes evolve imprinting because they have effects on relatives.

9) Do you have an opinion in regards to the data emerging from the HapMap project which implies a great deal of positive selection on the human genome within the last 10,000 years?

I have not given this much thought, but I would not be surprised because natural selection never stops.

10) If you could change one thing about your educational path, would that be?

I would have learnt more linear algebra as an undergraduate.