Peez

I will just add that natural selection can radically change the rate at which proteins evolve. If, for some reason, a change in cytochrome C was favoured in the primates, then we would expect to see greater change in the cytochrome C sequence. (I do not know if such natural selection occurred)

Peez

RufusAtticus

I've pulled CYC protien sequences from Genbank from 14 mammals and am using chicken as an outgroup.

I've found that the primate monophyly is well supported, but it does appear that probably the cyc in primates is/was under selection. That's because the primates have many more substitutions than other mammals. There also is an issue of long branch attraction because they cluster with kangaroos which also have many substitutions.

I'm now trying to do a phylogeny with nucleotide sequences.

Then I'm going to look at CYCB and CYCOx.

asgardhaven

Even if the nucleotide sequence works out to have the same proportion of relative changes as the amino acid sequence (ie few silent base changes that maintain the codon), is it really possible to establish a tree based on this one gene without looking also at the overlaying regulatory mechinisms, gene structure (ie introns), an genome position. Not to mention stuff like back mutation, differing selection pressures(?), etc.

Even if the creationist is correct about this particular gene isn't this perhaps just a case of shooting blind-folded in the general direction of a target until ten cases of rounds later you get a bullseye? We don't get to pick the one example that works for us and then ignore all the negative data. Why should they?

Just my two cents worth.

asgardhaven

Besides who said the tree for the gene is going to coincide with the tree for the whole organisms. Don't there need to be more data points to say anything even remotely like a sembalance of statistical accurracy as far as the ancestory of the species in question.

Ok so that's like an extra cents worth. Hope you guys don't mind.

lpetrich

There is a likely error in the mammalian common ancestor's reconstructed sequence.

#52 's value in that sequence is a, which agrees with the value in the rodents, but outside the rodents, it is either p or v, with p being the most probable.

And I agree that cytochrome c is not very good for mammalian phylogeny -- a faster-evolving molecule like hemoglobin may be more useful. In fact, I've recently made a huge collection of hemoglobin sequences, and I'm now trying to analyze it. I've selected those species that have both alpha and beta chain sequences available, so I can get two for the price of one. Also, the alpha and beta chains are recognizably related, offering automatic rooting.