Computing ddg_binding

[ajasja]

Hello,

I’m trying to use rosetta to compute changes in free energy of binding of a protein-protein complex upon mutation from wildtype.

On using the options listed at [http://www.rosettacommons.org/tiki/tiki-index.php?page=Commands+Tutorial#_7_1_calculate_ddG_for_binding|Commands Tutorial] along with the -mutlist option, my output (INT_OUT file) contains ddg_binding values for each of the mutations I listed. However, the contribution of the dunbrack energy to the ddg value is always zero.

Does this mean that the side chains are not minimized upon mutation?

I changed the interface weights in weights_manager.cc so that Wdun was non zero(I set it to 1), but the output still has dun energy to be zero. Is there some other reason the energies are zero? I have tried other variants in the input options including setting -repack, -softrep.

If the default behavior is not to minimize the side chains, is there any flag I can set to enable it? Also, If there is a set of recommended flags to enable while computing ddgs, that is different from the tiki page mentioned above, could someone please point me to it?

Many Thanks,
Hetu

[Anonymous]

There are two very important things you should be aware of while calculating ddG using analyze_interface protocol.

1) There is no minimization of the side chains at any stage. You can put -min_interface flag and that will do interface minimization while calculating ddG. But that will still keep your Edun = 0.

2) Edun is a one body energy term and what you see in your OUT file is dEdun( or delta Edun). If you look carefully, you will find that ddG includes only those terms which are two body (meaning both the partners are required). Since
ddG(binding) = dG(AB) – dG(A) – dG(
any Edun contribution in dG(AB) will be canceled in dG(A) or dG(.

Hope that helps.

—
Ramesh Jha

[Author]

Posts