Hacker News Reader

Comment by bh42 | Hacker News Reader

From: http://www.nature.com/news/2010/100420/full/news.2010.194.ht...

...the interbreeding happened about 60,000 years ago in the eastern Mediterranean and, more recently, about 45,000 years ago in eastern Asia. Those two events happened after the first H. sapiens had migrated out of Africa, says Long. His group didn't find evidence of interbreeding in the genomes of the modern African people included in the study.

I find this very surprising. We know that there was a genetic bottle neck when we left Africa. This is why today people from two different African ethnic groups can be MORE different then any two people from ANYWHERE in the rest of the world. An Inuit and an South American Indian are genetically more similar then a Touareg and a Lemba.

How could a separate species like the Neanderthals, who spent 200 thousand years in isolation in ice age Europe, contribute so little genetic diversity?

View on HN

How could a separate species like the Neanderthals, who spent 200 thousand years in isolation in ice age Europe, contribute so little genetic diversity?

A single successful interbreeding incident would contribute some odd genes to the local gene pool, but wouldn't contribute much genetic diversity once it spread out through the local gene pool. Remember that the data suggests exactly 2 individuals dropped into human populations that were locally in the tens of thousands, and globally in the multiple hundreds of thousands. It would contribute some odd genes to the pool of genetic diversity, but wouldn't move the bar very much.

For those who think that species barriers are some sort of absolute barrier to what is hypothesized here, it is worth considering the case of donkeys, horses and mules. Donkeys and horses are estimated to have diverged a couple of million years ago, but routinely can breed. The result are mules, of which over time we've bred tens of millions.. Mules are almost invariably infertile. However according to http://en.wikipedia.org/wiki/Mule#Fertility there have been more than 60 documented cases of female mules successfully bred by stallions. The foals would be mostly horse, and the result of that is an animal that is 3/4 horse which therefore is likely to be able to breed with other horses.

It is not unreasonably by comparison that there could be a constant level of interbreeding between us and Neanderthals, with many infertile offspring, and by chance on 2 occasions fertile offspring who have descendants today. (The overall number of crossbreeds would be lower than the number of mules, but their odds of fertility would be much higher than for mules.)

This would also suggest a reasonable chance that some hybrids bred back into the Neanderthal populations instead of ours.

Odds are high that those hybrids are now evolutionarily irrelevant.

Remember that the data suggests exactly 2 individuals dropped into human populations that were locally in the tens of thousands

I did not notice that, but it seems that's just what we have detected so far. If it can happen twice, isn't it very likely it would happen quite a few times?

As to species barriers, by definition a species is what is reproductively isolated.

Biology is not nearly as clean as say math, and so reproductive isolation is quite blurry. Some times A + B works and B + C works but A and C are reproductively isolated.

Often, A + B makes, C and C + (C or A or B) makes D, but D is infertile. Or E is infertile. And while A + B makes C and even D and E, the fact that the lines eventually terminate means A and B are reproductively isolated.

The foals would be mostly horse, and the result of that is an animal that is 3/4 horse which therefore is likely to be able to breed with other horses.

Not so.

It is true that once in a million a fertile mule is born, but that mule's offspring are not fertile, or its grandchildren are not fertile. Mules have been bred throughout history, but I know of no line of mostly donkey or horse hybrids with a fertile mule parent, or gand, or grand-grand-grand parent, or any mule linage.

and by chance on 2 occasions fertile offspring

But that in no way guarantees THEIR offspring with either Humans or Neanderthals (or other hybrids) would be fertile or that their grandchildren would be.

And if they were, then by definition us and Neanderthals are the same species.

And in that case I still don't understand how there could be so little genetic difference contributed.

For sweet zombie jesus' sake, we can identify the descendants of a Central Asian male who lived around the time of Genghis Khan over most of Eurasia. And successful interbreeding with a population who lived apart for 200 THOUSAND freaking years leaves only a few stray genes?

I had not looked enough into mule genetics. It is much odder than either of us had thought. See http://www.lovelongears.com/faq and search for "Fertile mules". It turns out that frequently a complete set of maternal genes get passed along. So there are recorded cases where a mule bred to a horse, and the result was a normal, fertile horse! There are also cases where she bred to a donkey, and those seem to come out either mule or 3/4 donkey.

In cat breeding female fertility is more resilient than male fertility. First generation females are fertile, but it is not until the third generation of breeding back to domestic cats that you get fertile males. If those two species are next to each other in the wild you could get some cross-breeding, but you'd never get significant gene transfer.

That said, the amount of gene transfer is more than this particular article had led me to think. See http://johnhawks.net/weblog/reviews/neandertals/neandertal_d... for a much better description of what was actually discovered. The estimate there is that non-Africans have 1-4% of our ancestry from Neanderthals.

In that case the answer to your question is that 4% injection from 200,000 years of straight genetic drift is less genetic diversity than 100% of very little mixing over 60,000 years.

Incidentally on comparing this research with tracing a given individual, we did that with a lot more data than we have here. We've only just sequenced a broad sweep of Neanderthal genes, then compared that to a handful of individuals and immediately got this strong genetic signal. As more data is acquired about what genetics are Neanderthal characteristics, I would expect the picture to get much clearer.

Thanks for that link!

It may not sound like a lot -- between 1 and 4 percent. But that's the equivalent of one great-great-great grandparent's DNA contribution.

That is quite a bit!

And from that same link:

Does this mean that Neandertals belong in our species, Homo sapiens?

Yes.

Reply on news.ycombinator.com