Prediction of protein structure

Artem Evdokimov AEVDOKIMOZ at
Tue Oct 9 19:35:51 EST 2001

Secondary structure prediction programs only can tell you the probability of
a particular stretch of the sequence to be found as helix, coil or sheet. As
such, they do not use templates - just weighting schemes and residue
correlation tables/chains etc. Software for prediction of tertiary structure
from scratch either does not work or needs a reliable sequence homology with
an existing model. I expect that software that actually predicts arbitrary
structure purely from sequence would win its author the Nobel, but I am not
holding my breath just yet :)


Oh yes, I am a crystallographer. All opinions are biased.

"Dominic-Luc Webb molmed" <domweb at> wrote in message
news:Pine.GSO.4.40.0110091639450.12593-100000 at
> You are interested in 2' and 3' structure prediction. The stuff I know
> about on Expasy uses templates for proteins with known structures, so
> is not a true prediction from scratch. If someone knows of software
> that does not require a template, that is indeed interesting.
> Cheers,
> Dominic

More information about the Proteins mailing list