I set up mr-rosetta which is running over 5 days and I guess it should be keep
running for very long time (ran it on a local machine with single processor.
260 residues, 1MOL/Asy). My question is where to look for the positive sign:
So far, the LLG scores from phaser are very low: 2.31-27.1;
From rebuild:
Run_1: ROSETTA SCORE: -1273.01;
Run_2: ROSETTA SCORE: -1427.29;
Run_3: ROSETTA SCORE: -1468.84
Run_4: in running
What would be the good ROSETTA SCORE will possible to lead to the solution?
How long do you think mr-rosetta could keep running?
Thanks!
Post Situation:
I can't answer your question without knowing a lot more about the system. Using the regular Rosetta scorefunction, about -2 units per residue is "good", so -520 would be an acceptable score. You're well better than that, which might be good. However, I have a feeling this code isn't using the standard Rosetta scorefunction.
The only true answer for "how long should it run" for a Monte Carlo protocol is "until you like the answer" - can you look at Run 1-3 and see if they fit your needs?
Thanks smlewis. The score I took was from Stage "GROUP_OF_ROSETTA_REBUILD_1":
rosetta_solution ROSETTA SCORE: -1483.00
But in the same log file, it also indicated:
component_solutions: None
So I was not so sure what is good or bad.
Since I was running in a local machine with single processor only, it will run forever (over a month). So I took one of the best over_best solution (pdb and mtz files), then run phenix.autoBuild, but the results was not good: R=0.49, Rfree=0.56. This is not going to be the right solution.
What should I do next?
Thanks!