Replies

Are you familiar with the concept of random error?

I think so. I believe that is what we are measuring. It is supposed to be random error which causes the divergence. The differences may be significant when matched with time. In any case the expectation value is calculated for the sequence differences. It is defined --

The Expect value (E) is a parameter that describes the number of hits one can "expect" to see just by chance when searching a database of a particular size. It decreases exponentially with the Score (S) that is assigned to a match between two sequences. Essentially, the E value describes the random background noise that exists for matches between sequences. For example, an E value of 1 assigned to a hit can be interpreted as meaning that in a database of the current size one might expect to see 1 match with a similar score simply by chance. This means that the lower the E-value, or the closer it is to "0" the more "significant" the match is. However, keep in mind that searches with short sequences, can be virtually indentical and have relatively high EValue. This is because the calculation of the E-value also takes into account the length of the Query sequence. This is because shorter sequences have a high probability of occuring in the database purely by chance.

If I read the values correctly, Danio(zebrafish) and Sparus (Gilthead Bream another fish) are represented by sequences which are more significant by 100 than Branchiostoma floridae and by 1 million than Branchiostoma lanceolatum.

OK. The evolutionary model (see p. 2158 of the Ciona reference) shows urochordates diverged first from cephalochordates and vertebrates; cephalochordates and vertebrates diverged from each other shortly thereafter. So, to first order, you would expect about the same distance between Ciona and Branchiostoma as between Ciona and Danio.

This is the analysis of ADH3 sequences from the paper of Canestro et. al. that I cited previously


Pair compared           time of divergence (fossil evidence)     ADH3 'evolutionary distance'
primates/rodents                  65-100 MY                                  0.06
Reptiles/mammals                 288?310                                     0.12
Fish/tetrapods                     400                                       0.23
Hagfish/gnathostomes               555                                       0.30
Cephalochordates/vertebrates       -                                         0.37

As I understand it, the 'evolutionary distance' is largely a function of the number of amino acid substitution rate. They then estimate from the correlation line that the Cephalochordates and vertebrates diverged around 690 MY ago. They didn't analyze the Ciona ADH3 (probably didn't have it when they wrote the paper), and so I can't tell you how it would fit in. Note, however, that quoting numbers generated by a canned program without a full understanding of the methods they've used, and the limitations of the results, is notoriously dangerous. As a chemist who does a lot of computation, I'm perpetually wary of it. I would strongly recommend looking at the data by hand, aligning the sequences, and then counting substitutions before drawing conclusions from BLAST; if you get about the same number it does, you're on safe ground. I started to do that last night with the Ciona/Branchistoma data with a consensus alignment of all ADH genes, and I got between 249 and 252 Ciona/Branchiostoma identities over a 340 amino acid consensus aligned seqeunce, and 264 and 192 identities between Ciona and two Danio genes.

Remember, standard theories claim that there was a massive increase in the genome around the beginning of the vertebrate line, and that one Ciona ADH3 gene gave rise to several ADH vertebrate genes. So, of the two I tested; one was a little more similar to Ciona, and one was a lot less similar. When you pick an ADH3 gene from a vertebrate, you're picking the protein from a family of daughter proteins that is most like the protein in Ciona. So, understandably, there is a little selection bias. The vertebrate ADH1s are also daughters of the Ciona ADH3 gene, and they're a lot less similar. ADHs are also an atypical family of enzymes; they have an enormous variability of substrates between different species.

One solution might be to pick a more conserved protein. I started with cytochrome b5 last night; it's short, has both highly conserved and variable regions, and is found in most every organism, and as far as I can recall retains the same function in almost all organisms. Unfortunately I have two papers to give at an ACS meeting in ten days time, and I don't have the time at the moment to explore this further. However, I promise I will return to it in early April.