In article <1885 at mailgzrz.tu-berlin.de>, macbeth at cs.tu-berlin.de (Andreas Pahl) writes:
>> I cloned and sequenced a protein from a gram-positive bacteria (streptomycete) and
> much more homology to yeast, fungi and mammalians than to E.coli. So my question
> is, what may be the reason for this?
>> The protein is highly conserved from human down to yeast (about 80% identity),
> but as fas as I know the split of eukaryotes and prokaryotes took earlier place
> than the split gram-positive and gram-negative bacteria.
>> Andreas Pahl email: macbeth at opal.cs.tu-berlin.de
It seems to me that (at least) three general assumptions are being
made when one compares sequences and uses their affinities as a
measure of the affinities of the organisms that bear them:
1. no lateral transfer has occurred;
2. the genes have evolved independently and at the same rate;
3. the genes are orthologous.
If any of these assumptions are violated, the affinities of the
sequences may give a false picture of the evolutionary history of the
hosts. Unfortunately, there is no _a priori_ way to test assumption
#1. Problems arising from assumption #2 can be addressed by building
a tree using a phylogenetic analysis, rather than drawing conclusions
from sequence identities alone (perhaps Andreas will find that the E.
coli gene is evolving at a higher rate than its eukaryotic and
gram-positive counterparts, which remain similar by virtue of their
conservatism). Assumption #3 can be tested by looking for a larger
family of genes related to the genes in question (perhaps Andreas will
find that his original E. coli sequence is *paralogous* to the others,
but that a second yet-to-be-discovered E. coli gene is *orthologous*
and fits much more comfortably with the known relationships of
eubacteria and eukaryotes).
Arlin Stoltzfus (arlin at ac.dal.ca)