Dear Dhanasekaran, There are many examples of molecular replacement failing even in cases where the model and target structure share 100% sequence identity. These examples illuminate several factors that MR search strategies are sensitive to, including percent sequence identity and related parameters like RMSD from target, the impact of loops, small domain movements etc.
So if you want to learn in more depth about what makes a "good model" and what makes or breaks an MR search, perhaps some googling combined with a bit of obsessive reading might help. I actually found a ton of thoroughly helpful articles and reviews for MR on the web without too much sweat. Best wishes, Raji On Thu, Sep 12, 2013 at 5:21 PM, Dhanasekaran Varudharasu < dhana...@gmail.com> wrote: > Dear crystallographers, > > I have solved a structure of a glucose > binding protein of CE4 family. When I try to solve the structure using the > same CE4 family enzyme as search model, it failed for many case. Finally, I > solved the with a same family enzyme used as search model. As soon as I > solved the structure, I superposed my final refined model with structures > of CE4 family enzymes which did not produce the good molecular replacement > solution for my enzyme. I found that all are having (Beta/alpha)7 fold and > superpose very well with my model. Whereas, some loop region are not > superpose very well. My doubt is why molecular replacement failed thought > over-all fold is same?. > > > -- > *Dhanasekaran Varudharasu* > Post-Doctoral Fellow > Department of Oral Biology > Rutgers school of Dental Medicine > Rutgers Biomedical and Health Sciences > Newark, NJ 07103 > USA > > > > -- Raji Edayathumangalam Instructor in Neurology, Harvard Medical School Research Associate, Brigham and Women's Hospital Visiting Research Scholar, Brandeis University
