Modelling Effect Of Mutation On Protein Structure

Question

[didymos]

Hi,

I am interested in structural changes in my protein due to mutation(s). I have WT pdb structure and I would like to model structure after single or multiple mutations . I would like to use something more fancy than just choosing from rotamer library. I was thinking about RosettaBackrub or Modeller or maybe there is another tool?

Thx for any suggestions!

Answer 1

[Gareth Morgan]

FoldX - http://foldx.crg.es/ - will calculate a new structure and have a go at estimating the stability change if you give it a pdb file and list of mutations. It's a downloadable program, which is fairly straightforward to use once you get started.

[didymos]

thanks I will give a try

[mikmaksi]

Be careful not to introduce too many mutations. The structure can be very unpredictable. In protein switches it takes just one mutation to switch the structure completely. You might try TINKER (http://dasher.wustl.edu/ffe/) as well. I think it'll do we with just "wiggling" the new sidechain into its best position.

[mikmaksi]

Be careful not to introduce too many mutations. The structure can be very unpredictable. In protein switches it takes just one mutation to switch the structure completely. You might try TINKER dasher.wustl.edu/ffe as well. I think it'll do well with just "wiggling" the new sidechain into its best position.

Answer 2

[João Rodrigues]

Modeller isn't likely to give you any hints. Since it's homology modelling, the entire protein, bar the mutation(s), is going to retain exactly the same structural features. Rosetta Backrub does seem a much better alternative for this.

If you can, get your hands on a MD package like GROMACS and run a moderately long simulation, depending on the size of your protein, so you can derive some conclusions out of it, although you won't see any spectacular unfolding or rearrangement. This will take a bit more time (a few days running, another few analyzing the results) but will probably give you a much more detailed picture of the influence of the mutations in the dynamics of the protein.

In the case of protein-protein complexes, I would look at the latest rounds of CAPRI and decide on who to choose to model the complex.

[didymos]

thanks! I was thinking about gromacs, but I have to model 250 structures so it would be too computationally expensive...