February 07, 2013

Tianyuan autosomal data

I had checked a few times after the Tianyuan paper came out, and it seems that the data from it is now publicly available. The data appears to be in SRA format; see instructions on how to convert to SAM format in the link.

11 comments:

genetiker said...

I predict that a globe13 analysis will show Tianyuan to be made up primarily of the East_Asian, South_Asian, and Palaeo_African components, in decreasing order.

gnomon said...

A thing about Tianyuan is fully expected from my 2008 paper.
Huang, S. (2008) Ancient fossil specimens are genetically more distant to an outgroup than extant sister species are. Rivista di Biologia / Biology Forum, 101: 93-108.

The tianyuan paper showed that tianyuan is far more distant to the outgroup africans than either extant chinese or europeans do. Should not he be much more closer, a lot closer rather than the opposite, if he has 45k year less time to mutate, and had separated from Afican for only 10-20k years? Why is such glaring contradiction gets overlooked by everybody?

I have looked a few more of the numerous fossil sequences published since my 2008 paper. My conclusion is that all confirmed my old paper. You can try for example the denisovan mtDNA (FR695060.1 ) or whole genome and you will find that denisovan is more distant to chimpanzee than extant humans are. simple blastn will do.

Nothing in the existing theory can account for this but it is fully expected from our new hypothesis, as shown by the recent papers available for download at my lab website:
http://www.sklmg.edu.cn/articles_98.shtml?l=en-us

genetiker said...

The tianyuan paper showed that tianyuan is far more distant to the outgroup africans than either extant chinese or europeans do.

Nothing in the existing theory can account for this

You must have missed this part of the paper:

"As suggested by the nucleotide differences (Table 1), the maximum-likelihood tree (Fig. S6A) shows that the branch leading to the Tianyuan individual is long, due to its lower sequence quality."

Grey said...

"The tianyuan paper showed that tianyuan is far more distant to the outgroup africans than either extant chinese or europeans do. Should not he be much more closer, a lot closer rather than the opposite, if he has 45k year less time to mutate, and had separated from Afican for only 10-20k years?"

Maybe the Africans they would have been closer to were rubbed out by the Bantu expansion?

gnomon said...

You must have missed this part of the paper:

"As suggested by the nucleotide differences (Table 1), the maximum-likelihood tree (Fig. S6A) shows that the branch leading to the Tianyuan individual is long, due to its lower sequence quality."

Yes, I did miss it and thank you for bringing it up. But it did not surprise me at all. Similar excuses have been used before with little creativity. I said the following on a similar finding with Neanderthals in a post here in my blog of May 2010, http://www.thegoldengnomon.blogspot.com/search/label/Neandertal

"the paper found Neandertal to be 4-6 times more distant to chimpanzee than extant humans are, fully supporting my earlier paper on all known informative fossil sequences (my 2008 paper). These authors are masters of cherry picking like most in the field and downplay this finding by stating it is due to sequencing errors. If an unexpected result is due to such errors, can one have any confidence on their other results? "

Now let's calculate what kind of error rate it would have to be to explain the table 1 distance data of this latest PNAS paper. The average distance of Han to the five Africans is 30115, while Tianyuan is 32203 (2088 extra dist than Han). According to the latest expert opinin by Henn et al of Stanford "The great human expansion" just published in PNAS, the asian and african split about 40K years ago (fig 1 of that paper). Thus tianyuan should be just 15057.5 away from Africans. But the data show 17145 extra length. The total SNP analyzed is 86525 (supposedly high quality as emphasized by the authors). So the error rate is 19.8%. the calculation here assumed that all distances starts from zero, which may not be true. So the error rate estimation may be inprecise. Butnonetheless,it is going to be too high for any meaningful results. bottom line, if error is an issue to cause such conspicuous inconsistencies, any consistent results could also be due to errors. One cannot have it both ways.

My explanation for this is simple. Most of these fast evolving sequences are under natural selection. Those seq for adaptation 40K years ago were mostly different from those of today.

genetiker said...

According to the latest expert opinin by Henn et al of Stanford "The great human expansion" just published in PNAS, the asian and african split about 40K years ago (fig 1 of that paper).

No, it shows it as 50-60 ka. But all of the details of that figure are wrong anyway. For accurate timelines of prehistory, go here.

Thus tianyuan should be just 15057.5 away from Africans.

You seem to think that genetic distances are simply directly proportional to time. If that were the case, then all extant human races, and indeed all extant species, would be genetically equidistant to their common ancestors. But of course they're not, and of course that's because selection, drift, mutation, and gene flow have affected different species, and different human races, differently.

I said the following on a similar finding with Neanderthals in a post here in my blog of May 2010

I read your blog post. My advice to you is to abandon Weidenreich's idiocy, and embrace reality. You'll feel much better when you do.

But I did find one fleck of gold in all the mud:

"Finally, the paper notes the striking finding that equal intermixing of Neandertals with Europeans and Asians did not translate into equal physical resemblance of Europeans and Asians to the Neandertals."

That's because incipient Mongoloids didn't intermix with Neanderthals. When one mentally juxtaposes images of Caucasoid heads, Mongoloid heads, Neanderthal skulls, and the Dali skull, it becomes obvious what happened.

gnomon said...

Yes, there is little certainty in the separation time of the human groups. Anywhere from 20-70K is possible, including any other time outside that range. Because the assumptions for these estimates are oversimplification and could be and are in fact wrong.

Regardless of the time used for error rate estimation, our analysis already disapproved the sequencing error excuse. We have shown that slow evolving sequences do show that Tianyuan is closer to the non C and nonD (Y chr haplotype) male JPT, CHB, CEU, JPT C/D type male, and YRI, in the order of high similarity to low. Tianyuan is closer to the outgroug CEU or YRI than a CHB or JPT is. The results are fully consistent with a multiregional perspective (manuscript to be submitted to ArXiv soon). And there are no ad hoc epicycles necessary to explain any of the observations, including the European and Asian mixed phenotypes of the tianyuan fossil. Everything is as normal and perfect as could be expected.

I must give Da Paabo great credit for his superb quality in fossil DNA sequencing. But his interpretation of his own data is a whole different story. But we should not blame him for that and he was merely going with the flow as demanded the law of natural selection.

genetiker said...

globe13 results

truth said...

Genetiker that looks suspicious...

genetiker said...

Genetiker that looks suspicious...

I invite you to do it yourself.

gnomon said...

Phylogenetic analysis of Tianyuan using informative DNAs