Dear netters:
The problem of prediction of transmembrane protein topology does not
just in location of helices by elucidation of hydrophobic regions and in
their subsequent arrangement according to bacteriorhodopsin or rhodopsin
hypothetical structures. You will still loose interhelical loops and still
will assume that all transmembrane proteins are similar in their 3D
structures. Few available commercial (and not) programs cannot go further
than these assumptions.
At the same time, we believe we have developed a computational
approach to deal with this problem. In short, our approach to this problem
includes three main steps: to locate transmembrane helical fragments in
the sequence; to pack these helices together; and to restore interhelical
loops. At all these steps, we are using novel computational techniques and
software, which are not available commercially. All three techniques taken
together were applied to prediction of 3D structure of the L subunit of
the photosynthetic reaction center, the only transmembrane protein with
the known X-ray structure. The predicted general 3D pattern for this
transmembrane protein was very similar to that revealed by X-ray studies.
Thus, our 3D models for transmembrane receptors are customized for each
particular receptor; they are not based on bacteriorhodopsin or rhodopsin;
and, they include interhelical loops. To the best of our knowledge, this
level of accuracy and reliability cannot be achieved by any other approach.
At present, our programs are not available commercially, nor they are
distributed through any network. However, if somebody is involved in a
project, which might be helped by 3D models of transmembrane receptors, this
person is welcomed to contact me to discuss possibilities of collaboration
on the contract basis. Please e-mail me personally, not to the group.
Gregory V. Nikiforovich
