Hi everyone, I am using refine.py to refine a structure in Xplor-NIH and submitted the top 20 structures to the PSVS server for validation. The models came back with a high MolProbity clash score (~51) indicating VDW clashes.
I checked the output of script.out for all the models and there is this error statement at the end Powell::step: irregular exit: Line search abandoned: gradient may be incorrect *-- POWELL ------ step= 25 --- stepsize= 0.01000 --- energy evals= 41 -* | E(poten)= -181.7321850 grad= 4.1760241 ANGL= 120.7184528 | | BOND= 20.0459054 CDIH= 43.1371207 HBDA= 0.0000000 | | IMPR= 89.3391424 RAMA= -791.2350898 VDW= 140.4722961 | | Vgyr= 0.0836804 jcoup= 0.0000000 noe= 122.4972026 | | rdc= 73.2091044 | *------------------------------------------------------------------------------* .xplorInput.30541(378): loopInfo.writeStructure(potList) .xplorInput.30541(379): pass Structures were still created and they have few violations and good rmsds. I wonder if any cartesian minimization is being done to my structures or if this error message indicates that the program abandoned minimization but still created structures. Is there any way to increase the number of steps in the final minimization and does anyone have an example script ( I am new to xplor-nih and python). Thanks, Ed
