Hi , I was just wondering if it’s realistic enough to get the docked protein’s RMSD value close to 0 when compared to Co-crystallized structure? I know that will be the ideal structure but practically is it achievable? If not, what are the RMSD value that are consider good enough for further analysis ?
Hi, never tried this myself, but in principle, this should be achievable for the backbone RMSD - if one performs bound docking with unambiguous restraints. Also, this is likely to depend on the experimental structure itself - it may be more or less optimised, it may have not-great resolution, etc.
Thus, sometimes people use an energy-minimised experimental structure as a reference instead of the original coordinates that have been deposited on PDB.
For real-life protein-protein docking, one option is to use CAPRI metrics:
- acceptable model: i-RMSD < 4Å or l-RMSD < 10Å and Fnat > 0.1 (0.23 < DOCKQ < 0.49)
- medium quality model: i-RMSD < 2Å or l-RMSD < 5Å and Fnat > 0.3 (0.49 < DOCKQ < 0.8)
- high quality model: i-RMSD < 1Å or l-RMSD < 1Å and Fnat > 0.5 (DOCKQ > 0.8)
Good day. Thanks for the reply, yes unambiguous restraint may lead to a lower RMSD. However, my group leader may argue that over restraining the protein interaction may not be natural enough.
Anyway, is there a way to determine the I-RMSD or FNAT ? As so far I had been using ChimeraX as my molecular viewer and the RMSD there I suspect it to be backbone RMSD
You could use the CAPRI-Q resources for CAPRI metrics assessment:
The source code is freely available from CMSB-public / CAPRI-Q · GitLab
and its web interface through the Dockground resource at Dockground - Assessment
Thanks Prof, you are a lifesaver 