From my alignment results, there is a consecutive sequence (around 32 amino acids) which is present in my target but not in my template. After I modeled my protein with MODELLER, that sequence resulted as a long strand which do not pack together with my protein. I felt very unusual about this hanging long strand. And since I am a beginner for protein modeling, I really have no idea about this. I have tried to find the solution, there is a saying that the tail sequence can be omitted. Is that true that I can edit the alignment by removing the sequence for this hanging strand? Or anything that I can do for that? Thank you.
Yes, you don't want to include these parts in the model - in experimental structures, you also frequently find that there are parts which are missing in the structure, because there is no information available on their structure. You can either remove these parts from the initial sequence and redo the modelling, or you can delete these residues from the pdb file. Of course, you'll mention in the description that "residues x-y could not be modelled". Sometimes you can get some information on these parts using multiple modelling templates or ab-initio folding methods, but unless you absolutely need a complete model, it is better to leave out those parts of the model for which you have no reliable prediction.
Yes, you don't want to include these parts in the model - in experimental structures, you also frequently find that there are parts which are missing in the structure, because there is no information available on their structure. You can either remove these parts from the initial sequence and redo the modelling, or you can delete these residues from the pdb file. Of course, you'll mention in the description that "residues x-y could not be modelled". Sometimes you can get some information on these parts using multiple modelling templates or ab-initio folding methods, but unless you absolutely need a complete model, it is better to leave out those parts of the model for which you have no reliable prediction.
Should I just remove the sequence and run the modeling process again or I need to edit the alignment through some tools?
I would see whether that 32 aa have any matches in other structures in the pdb and if yes, would go for a multi template modelling. In case nothing is available, either delete it in the input query sequence, or the alignment or just delete the residues from the final model and run a few cycles of energy minimization.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics