After scanning the literature for a few more days, I have
come to the conclusion that the patterns of mutation I am observing
have not been adequately studied by other people yet. So I want
to study them myself and publish the results.
What I need is software that will take a multiple sequence
alignment and tabulate the mutations observed. It is not a trivial
problem because if we take 3 sequences:
A gatctgtcagct
B gatccgtcaggt
C gttctctcaggt
muts * ** *
It looks like C mutated a -> T at postion 2, B mutated t -> c at
position 5, and so on...
However it could also be true that the common ancestor of A
and B mutated from t to a at postion 2 after that AB ancestor had
split from the ancestor of C. Or both A and B could have suffered
identical point mutations.
I have 2,700 bases of protein-encoding DNA from human, rat
and hamster. I count 470 total mutations. 88 in rat vs human-hamster
consensus. 117 in hamster vs human-rat consesus. 165 in human
vs hamster-rat consensus. 100 where all three are different.
Counting how many appear to be a G -> A mutation, how many
appear to be a G -> T mutation etc by hand I see some striking
features. But my eyes went berzerk on these 2,700 bp and I now
need to look at a few more genes to make a statistical arguement
for my findings.
I am starting to write some spreadsheet macros that will
do the counting for me, but if any of you know of existing software
to do this, please let me know. I'm sure it will take me a few days
to get a good spreadsheet working.
--
********************************************************************
* Brian Foley * If we knew what we were doing *
* Molecular Genetics Dept. * it wouldn't be called research *
* University of Vermont * *
********************************************************************
