Hi Ben:

I already modeled in the past, using modeller, this trimer from chicken to humans.

Now, from new, low-resolution x-ray diffraction data for that trimer in complex with a small molecule, I merely want to reconstruct the missing side chains (many of them) while omitting the missing residues (these are at the beginning and final portion, i.e. protruding from the pore into the intracellular). The lacking residues are many and for a flexible region that resists any modeling, and I am not interested in that part of the transmembrane protein.

My question is whether <http://salilab.org/modeller/wiki/Missing%20residues> can be adapted to do that for the trimer. Unless I did something wrong, I got the answer "no missing residue', in fact I had eliminated the flexible portions. In other words, what I would like to do, is reconstructing the missing atoms (side chains) in the context of the trimer, not each chain at a time.

Thanks
francesco pietra

On Jun 23, 2013, at 7:00 PM, Ben Webb <ben@salilab.org> wrote:

> On 6/22/13 10:17 PM, Francesco Pietra wrote:
>> Modeling a trimer (all three chains present, though missing residues and
>> atoms) with chimera and modeller installed locally.
>>
>> The default is modeling each chain, one at a time. The chains are
>> identical in sequence but the X-ray structure shows different localized
>> conformations. Question: is modeller modeling by taking the other two
>> chains into account?
>

> The Modeller mailing list would be more appropriate for such a question (unless it's specific to the Chimera Modeller interface). If you want to model a trimer using a trimer template, you don't need to do anything special (just include the entire sequence of each trimer, with '/' chain break characters where appropriate). Modeller will consider all interatomic distances. (If the Chimera Modeller interface isn't doing this, the Chimera people may want to comment further.)
>
> If you want Modeller to model (e.g.) chain A using structural information from each of chains A, B, and C, that's just multiple template modeling (using A, B, and C as templates for A).
>
> Ben