Most annoying evolutionary genetics misconception

I happened into a post on Uncertain Principles soliciting candidates for "Most annoying misconception about your field". The choice there was the nature of the Heisenberg Uncertainty Principle.

RPM at Evolgen posted an annoying genetics misconception, namely:

the idea that by sequencing a genome, we are "decoding" it

Personally, I think this would be a great topic for a meetings panel. I can think of about fifty really annoying misconceptions. The nature of these two seems especially science-fiction-y -- the sort of thing that people generally knowledgeable about science sort of understand to be true or meaningful.

On that level, I would nominate the one-gene, one-trait fallacy -- the idea that every trait is encoded by one gene. From this we have the popular idea that you can discover a gene "for" something. Of course, the usual case is that you discover a gene with a particularly bad allele in some people with a genetic disorder. But that generally gives little clue about what the gene does, or what its evolutionary history might be.

But that isn't actually the one that annoys the most.

Instead, I am most annoyed by the idea that genetic divergence and species divergence are the same. That fallacy underlies the long literature on the 5-million-year divergence date for chimpanzees and humans.

Set aside for a moment the fact that the species divergence must be at least old enough to encompass the earliest fossil hominids (now, around 6 million years old or so). The real problem is that two newly-isolated populations are not genetically uniform -- each of them has variation and the two overlap to some extent. The date of genetic divergence between alleles taken from each of the populations at the time of their isolation is a function of the demographic and selection histories in their common ancestor population.

So genetic divergence must always precede species divergence. That period of time probably ranges from a few hundred thousand to several million years before the time of species divergence.

It's an important problem in the human-chimpanzee case, since the 5-million-year genetic divergence would correspond to a younger species divergence. Genetics are even more out of tune with the fossils than most people usually assume. (The problem is likely in accurately estimating the rate of genetic changes.)

And it's always annoying to see that 5-million-year figure repeated endlessly when, not only does it have to be wrong, it shouldn't have even been applicable in the first place.