denovo modelling on a virus-coat protein dimer

[Anonymous]

I am new to denovo modelling so your help will be greatly appreciated. 

 

I have a density map for a coat protein dimer of a virus (no previous structure of same or similar virus is known so homology is not an option). The two chains seem to swap domain so i cannot segment the dimer further. Rosetta seems incapable to handling a dimer for denovo modelling (each chain is 410 AA long) so i used phenix.map_to_model to get a partial map. 

How do i use this partial map to invoke RosettaCM for homology modelling to get a complete model ? How should i generate the alignment file when there are two chains of the same sequence but phenix has build (very) slightly different region of each of the dimers

 

Thanks in advance, Aseem

[Anonymous]

I would suggest looking at:

https://dimaiolab.ipd.uw.edu/software/

for information on how to use rosettaCM in this situation, but I would use your phenix model to generate a symmdef file as described (if possible) and then run Either rosettaCM or RosettaES to complete loops. (I would run rosettaCM first, and then rosettaES on the parts you don’t like)

 

Once you have your symmdef file you should be able to provide CM with a single sequence, which will be modeled symmetrically.

 

If you are really stuck, I would suggest contacting Frank or me directly and we can maybe collaborate on a solution. 

[Author]

Posts