I am working on a transmembrane protein. It is an efflux pump which belongs to MFS super family. I want to predict its structure and conduct docking studies. What are different methods for predicting the structure of this particular protein as it has only 34% homology?
Hi Yugendran,
I have no practical experience with such an analysis, but Rosetta could work fine. Or you can try to predict a contact map which can reduce possible conformations.
https://www.rosettacommons.org/software
The best way is to use Ab Initio method. You may try I-TASSER as well. It gives the good results with proteins having lower percentage of homology.
There is also Phyre2 as well. This gives a lot of additional computational analysis as well such as multiple sequence alignments, secondary structure prediction, motif searches, predicted binding sites etc..... You can find it here http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id=index
All you need to do is input your sequence
Thank you so much all for the suggestion.
I have predicted structures using Phyre2. The I-Tasser reult is not available yet. I validated the structure generated by Phyre2 and the ramachandran plot has only 84% of aa in allowed region and 2.5% aa in disallowed region. Is this a good structure. Will the docking studies carried using the structure be correct? Is there any means to improve the Ramachandra plot score.
Try ITASSER and You may Raptor X
http://zhanglab.ccmb.med.umich.edu/I-TASSER/
If the percent identity ranges from 27-38% and you have almost complete coverage, you can try SWISS-MODEL which bases predictions on homologous structures. Failing that, you could try tools that use fold recognition or Ab Initio folding like RaptorX or I-TASSER.
I would generate a bunch of models by at least 3-4 publicly available servers that proved to be good modeling tools in CASP: http://predictioncenter.org/index.cgi?page=links. Modeller and Swiss-model in this identity range should be also fine. After that you can do some clustering quality assessment (see our latest CASP10 quality assessment paper for servers).
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics