Predicting Effect Of Mutations In Terms Of Activity

Question

[Pappu]

I have a membrane protein (~400 aa) in which >90% of the mutations result in the loss of activity. I was trying to use SVM to predict the effect of mutation. However it predicts loss of activity in all cases. Let me know if it is possible to use PSSM matrices along with SVM or other tools to better predict the effect of mutation.

Answer 1

[AsoInfo]

You can try using project HOPE.

http://www.cmbi.ru.nl/hope/home

[Pappu]

I don't think homology model will work because it is a membrane protein without any structural homolog. I have already tried homology/ab-initio modeling. But how can one validate such model most of the mutations are causing loss of activity?

[AsoInfo]

You want to assess the biochemical effects of mutation in protein or you want to validate a protein structure using different tools?

[Pappu]

I want to assess the biochemical effects of mutation in protein.