I have a protein that binds specific short dsDNAs. I have docking results from both HADDOCK, Lightdock and RosETTAFold2NA, but not sure which to go for in GROMACS and what force field. I have reasons for the suspicion that the protein binding to the DNA causes the protein to change conformation and the DNA strands separating. How much does the ability of GROMACS runs leading to this depend on the quality of the docked model, and how long at minimum would be expected for such a conformational change?
Hi,
I’m afraid it’s very difficult to give any general answers, as also indicated by the lack of replies. One thing I can say is that the quality of the docked model will affect the results. Starting from a wrong, or unrealistic, configuration may make it improbable to reach a correct result.
Regarding the times for configurational changes, it is very system dependent. I think you would know best how long such a configurational change is expected to take for your system. It can be anywhere from picoseconds (not probable) to seconds or even minutes. MD simulations can routinely access time scales of microseconds, but even that may require significant hardware resources.
Using the parmbsc1 force field sounds reasonable, but, as always, the results should probably be compared to some kind of experimental data to be on the safe side.