I suspect what your colleague wants is not a whole slew of homology models,
but for you to take a look at variable residues, see what they are doing in
the structure, and make comments like
The tyrosine is H-bonding the ligand, so mutating to phenylalanine may
weaken the binding or change specificity, the Ile is in a hydrophobic patch
and mutating to Leu probably won't have much effect, or the Ile is in a
hydrophobic patch at the bottom of the active site cleft, so mutating to Ala
may allow a larger substrate to be accommodated.
If (s)he doesn't feel coonfident doing that for himself, he probably
wouldn't feel confident drawing conclusions from the homology
models either.
Simon Kolstoe wrote:
> Dear ccp4bb,
>
> One of my colleagues is interested in how a certain protein differs between species. He's done a blast search, collected all the aligned sequences, and "emailed them to the crystallographer to tell him the implications of the sequence changes".
>
> Although I am not at all confident that I can predict any implications based upon sequence differences, I thought I could at least have a try by mapping the different species sequences onto the existing structure and then regularising it just to see what happens (and then perhaps looking at buried surface area, electrostatics, subunit interfaces etc.).
>
> I know the program chainsaw can mutate the sequence based on an alignment, however I can't stop it pruning non-conserved residues. Does anyone know of another program that I could use to do this step?
>
> Similarly if anyone knows other software tools/methods I could use to try and work out the implications of sequence changes I would be grateful for advice.
>
> Thanks,
>
> Simon
>
|