Hey guys, I counterpart with problems as topic saying:
First, I want to use others computational results as reference to optimized docking option, so I have to relax them into same scorefunction, meanwhile relax the inputs which generate from different methods for better results and avoid disulfide bonds error due to different scorefunction, right?
however, as I run rosetta_scripts.linuxgccrelease for mpfastrelax on my identical subunits multi-chain, error occurs as mpfastrelax_log.txt describe, seems the beginning mover cause the problem(flag as mpfastrelax file),but when I use mprelax for ensemble (flag relax_for_ensemble), it complete the work. I avoid use relaxed model to generate initiate model and failed during prepack (flag prepack_ensemble) as log in prepack_log.txt
Second, as look at in rigid dock us franklin2019 scorefunction, it seems only output the interface residues score about i_sc etc. but still little terms without explain in reference or tutoria, like dock_ens_conf and st_rmsd, and does single protein with relax in franklin2019 will output score terms correctly? could I relax all the inputs in same scorefunction? And what about the I_sc terms range in the membrane environment? Besides mpframework_docking_fa_2015 output terms CAPRI_rank value to 1.84467e+19, does it refer to CAPRI competition scoring criteria and correct?
Third is the docking ligand restrict,mpdocksetup method will split a subunit of multi chain complex and dock partner in the place, so I generate initiate model manually and I have already point out the partner as ABCD_EFGH (flag MPdock),but finally the chain B still disperse from multi-chain complex and act as docking ligand, move around as picture(only unexcepted ligand chain b show in different color ,other partner did not move, attached initiate model and several results), seems I have to convert multi-chain into single chian? But the gap convert into super long loop must be extraordinary.
and suggest might be helpful, thanks.
lluoto