I have a sequence that I would like to model by threading onto each
subunit of a homo-tetramer of a related protein (crystal structure). It
doesn't work to thread each subunit individually and then 'recombine' the
homodimer as a single PDB because the interfaces suffer from clashes and
overlaps. Does anyone know of a server that can peform multimer threading
based on a sequence and a multimeric crystal structure?
when something doesn't work you always have to wonder why. in this case, i
think you're using the wrong method and asking the wrong question. threading
is a method for fold recognition, which in this case seems unnecessary since
you know the fold apparently. the alignment (and, hence, model! a model is a
3D representation of your alignment, after all) you get from fold-recognition
methods is often poor.
so, you need to generate a reasonable alignment first, then use comparative
modelling techniques (e.g., andrej sali's Modeller program) to generate a
model for the monomer, then assemble the homo-tetramer. now you have to
*validate* your alignment by checking if the interactions in the interface
make sense or not. there will probably be clashes that can be relieved by
choosing different rotamers. there may also be clashes that cannot easily be
resolved. in those places you need to adjust your alignment - perhaps a
register shift of +1 or -1 residue along a beta-strand suddenly results in a
sensible interface - who knows.
with the improved alignment you iterate the whole process. once you have a
reasonable tetramer model, you can put in better rotamers and do some MM or MD
to resolve clashes and relax the model. then run it through whatcheck or
molprobity to find places where the model can be improved further and repeat.
or you grow crystals.
--dvd
******************************************************************
Gerard J. Kleywegt
[Research Fellow of the Royal Swedish Academy of Sciences]
Dept. of Cell & Molecular Biology University of Uppsala
Biomedical Centre Box 596
SE-751 24 Uppsala SWEDEN
http://xray.bmc.uu.se/gerard/ mailto:[EMAIL PROTECTED]
******************************************************************
The opinions in this message are fictional. Any similarity
to actual opinions, living or dead, is purely coincidental.
******************************************************************
