Dear colleagues,
I need to evaluate the effects of mutations in all positions of query protein sequence (protein sequence randomization) in order to find the most destabilizing mutations. Doing it manually one-by-one is impractical. Is there any software for such task? I have both sequence and structure of the protein.
Thanks in advance for any suggestions.
The best available option currently is a web-server DUET (http://bleoberis.bioc.cam.ac.uk/duet/stability) that allows to asses the impact of all possible substitutions on protein stability in each position. But it takes only one position per time, so it's too long to analyse the whole protein.
You can model tour protein and check protein folding in silico by using tools in Expasy
Dear Claudia,
thanks for your suggestion. Unfortunately, it won't work in my case, since all these tools allow us to analyze the impact of single mutations on the protein structure, while I need to evaluate the impact of all possible substitutions in all positions of query protein sequence. So, I'm looking for high-troughput software that can assess the impact of hundreds of substitutions per one run. According to literature PROSA 2003 software is capable of doing this. Unfortunately, it doesn't work on my PC because of some probelms with license file( So I hope to find some alternatives.
The PoPMusic webserver will consider all possible single mutations for a single structure and is known to be fairly accurate. It takes only a few seconds per structure.
http://dezyme.com/
http://www.biomedcentral.com/1471-2105/12/151
Dear Jeffrey,
Thank You very much for the answer! This tool is exactly what I was looking for.
Dear Sergii,
there is plethora of tools for your desired purpose. PoPMuSiC definitely belongs to the most powerful ones as it adopts knowledge-based potential (which is less prone to overfit the decision model). According my experiences, the best results can be achieved by Rosetta (energy function - hybrid of empirical and physical-based potential) and FoldX (energy function - empirical potential). These methods are computationally more demanding as Rosetta could take about 1-5 hours per each mutation and FoldX about 1-5 minutes. Their installation and settings is also more complex but if you need accurate result you should consider it.
Finally I mention several comparisons of prediction tools:
Potapov, 2009:
http://www.ncbi.nlm.nih.gov/pubmed/19561092
Khan, 2010:
http://www.ncbi.nlm.nih.gov/pubmed/20232415
Thilgen, 2012:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3483175/
Hope it helps!
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools