Hi, I've recently built a homology model and have two questions that I can't find answers to in the documentation.
Question 1: My template structure has hetero atoms in it. Can I 1) make these carry over into the homology model and 2) use them as restraints in building the homology model. (The hetero atoms are in contact with the template structure just as I expect them to be with the homology model. I would like this to be realized in the homology model.)
Question 2: This is more a request for advice on analyzing my homology model. What is the best way to see if it makes sense? For instance, I would like to know if the core is suspiciously populated with energetically unfavorable residues, if one face of the protein is too hydrophobic, implying that it ought to be in hydrophobic contact with (in my case) the other half of the protein for which I cannot built a homology model, etc.
I could do something like color hydrophobics and hydrophilics different colors and visually inspect the model, but I'm interested in something a little bit more convincing and principled. What would you do to convince yourself that the protein could be physically realizable?
Thanks for any suggestions, Christian Barrett
Christian Barrett wrote: > > Hi, > I've recently built a homology model and have two questions > that I can't find answers to in the documentation. > > Question 1: > My template structure has hetero atoms in it. Can I 1) make these > carry over into the homology model and 2) use them as restraints in > building the homology model. (The hetero atoms are in contact with > the template structure just as I expect them to be with the > homology model. I would like this to be realized in the homology > model.)
Yes to both 1 and 2, probably. Please see Question 16 in the FAQ chapter in modeller-4 manual (BLOCK residues, indicated by '.' in the alignment); also at the beginning of the 2nd Chapter.
> Question 2: > This is more a request for advice on analyzing my homology model. > What is the best way to see if it makes sense? For instance, > I would like to know if the core is suspiciously populated with > energetically unfavorable residues, if one face of the protein is > too hydrophobic, implying that it ought to be in hydrophobic contact with > (in my case) the other half of the protein for which I cannot built a > homology model, etc.
We generally run Procheck (J. Thornton) and ProsaII (M Sippl). Most of the display requests you have can be satisfied by any decent structure display program (LOOK, InsightII, QUANTA, SYBYL, OXFORD MOLECULAR, and all sorts of academic programs).
> > I could do something like color hydrophobics and hydrophilics different > colors and visually inspect the model, but I'm interested in something > a little bit more convincing and principled. What would you do to > convince yourself that the protein could be physically realizable?
In my opinion, an energy profile and a Z-score from ProsaII (or Procyon or Profit) are really useful: http://lore.came.sbg.ac.at/. Also Gert Vriend's WHAT-CHECK is good: http://swift.EMBL-HEIDELBERG.DE/
Best, Andrej
> > Thanks for any suggestions, > Christian Barrett
-- Andrej Sali, Assistant Professor The Rockefeller University, 1230 York Avenue, New York, NY 10021-6399 tel +1 212 327 7550; lab +1 212 327 7206 ; fax +1 212 327 7540 e-mail sali@rockvax.rockefeller.edu; http://salilab.org