Entering edit mode
6.5 years ago
Arindam Ghosh
▴
510
I am trying to model a protein of length 438aa. It has ~88% identity and ~80% query coverage with existing PDB structures. I had initially used template based modelling using PHYRE2 and SWISS-MODEL. Both of them could not model between residues 1 - 68. The generated model was good. I then went to model via ITASSER. Even here the part 1-68 though modelled but as coils and exists as outliers in Ramachandran plot. The rest of the portion is almost same to the models generated by PHYRE2 and SWISS-MODEL.
- Please advice on how to proceed for MDS and Docking.
- Should I consider the portion that has been modelled correctly only?
It depends what you intend to do with it, and what that region represents.
Can you use anything known about existing structures to work out whether the first 68 amino acids do much that is useful or relevant? If not, and your active sites etc. are elsewhere in the model, then you’re probably OK to proceed cautiously. Simulated models are rarely 100% accurate, unless the crystal structure has basically been solved at >>90% identity.
Thank you! I had aligned my structure to available PDB structures and the modelled region completely aligns with RMS < 2. I did a cross-check on the protein and found it probably contains an initial target sequence (from TARGETP:http://www.cbs.dtu.dk/services/TargetP/)and from literature survey it was understood that it may have post-transnational modification. Now for going further I guess I have to remove the signal sequence and then model the protein. But how do I take into account if any PTM occurs?
To my knowledge there is no way to model PTMs with ITASSER and Phyre, but I work with bacterial sequences mostly so it's not usually an issue for me. If there is specialist software out there for refining models for PTMs, google may tell you - or maybe someone else on the forum knows more than me.e