My objective is to model a protein for which the experimental structure is not known. I performed homology modeling using Modeller and performed Energy Minimisation (for 1ns) using GROMACS to improve the quality of the model. I have got 10000 pdb files from md.trr file using 'trjconv' command.
Now, the last pdb (that is, 10000th pdb file) has got 3 outliers when checked using Ramachandran plot. Is there any way, I can pick a better structure out of these 10000 structures generated, using any Bioinformatics tools?