|
Comparison of Phylogenetic
trees based on sequence alignments and structure alignments
-
There is a general need to
have some parameters to aid comparisons of sequence and structural
alignment. Phylogenetic trees generated from sequence and structure
alignments might be a good choice for evaluation of such alignments.
-
We have created a flow-chart
for the procedure that offers convinient and indicative evaluation
of phylogenetic trees generated by two different alignments. This
in turn, provides a way to evaluate a correctness of the two types
of alignments. This procedure uses SCOP Serine Protease classification,
Web available ClustalW algorithm for sequence alignments and phylogenetic
tree generation and AnSuei Yang's program: PRISM for the Multiple
Structure Alignments.
|
We have demonstrated (using Phylogenetic trees based on
sequence and structure alignments) that structure alignement definitly
evaluates better protein phylogenetic relation than sequence based
analysis only. For this purpose we did not use "fine tuning" or
iterative multiple sequence alignment since these would introduce
obliteration of the difference of the two procedures!(or at least
strong approximation between two procedures, by introducing
"structural knowledge" to sequence only information).
|
|
It is obvious that sequence alignment is biased toward linear code
superpositon (by definition). This often blures the information on
evolutionary correlation because the latter one is much better
contained within 3D fold conservation. Full function of the certain
protein class, or altered functionality among protein subclasses is a
product of 3D conservation of the specific fold, rather than
conservation of patches of sequences.
|
|
