Generating Constraints File and Pre-Minimization Problems with ddG_Monomer Application -
Hello,
Recently, I have had some trouble preparing my input files for the High Resolution approach for the ddG_monomer application using Rosetta 3.12. After running the pre-minimization on FoldX repaired and Rosetta renumbered PDBs I am unable to convert the minimization log file to a constraint file. While the convert_to_cst_file.sh command works without an error output, the resulting file is empty. Some of the PDB structures that I am using, specified in a lst file, are dimers, and I am unsure about how this could impact the pre-minimization run. I have included outputs from the mincst.log file that was generated and also want to note that I am performing the pre-minimization through a computing cluster as a job to utilize more resources than my computer can manage, (1) could the issue have something to do with using multiple nodes and cores for a job?
(2) What can I do to resolve this occurance of the empty constraint file? (3) Is there an issue with the pre-minimization run, as each pdb from the lst in the mincst.log file ends with multiple error messages, despite generating constraint information throughout the run?
Please let me know if there is other information needed to understand this issue. I am unable to attach the mincst.log file itself due to its large size.
Thank you!
Example of errors during pre-minimization as written into mincst.log file(each PDB structure (5 total) specified in the lst file has these errors at end of their turn):
apps.public.ddg.minimize_with_cst: running another iteration of minimization. difference is: -1431.95
core.optimization.LineMinimizer: [ ERROR ] Inaccurate G! step= 4.76837e-09 Deriv= -431.913 Finite Diff= 12.4594
core.optimization.LineMinimizer: [ ERROR ] Inaccurate G! step= 5.62989e-09 Deriv= -194.02 Finite Diff= 20.7298
core.optimization.LineMinimizer: [ ERROR ] Inaccurate G! step= 2.66702e-07 Deriv= -818.329 Finite Diff= 2.1728
core.optimization.Minimizer: Line search failed even after resetting Hessian; aborting at iter#310
apps.public.ddg.minimize_with_cst: running another iteration of minimization. difference is: -1.70451
core.optimization.LineMinimizer: [ ERROR ] Inaccurate G! step= 3.8147e-08 Deriv= -821.016 Finite Diff= -1.58947
core.optimization.LineMinimizer: [ ERROR ] Inaccurate G! step= 2.66265e-07 Deriv= -821.016 Finite Diff= -0.514863
core.optimization.Minimizer: Line search failed even after resetting Hessian; aborting at iter#1
Command used for pre-minimization:
/mnt/home/dolgikhb/Rosetta/rosetta_bin_linux_2020.08.61146_bundle/main/source/bin/minimize_with_cst.static.linuxgccrelease -in:file:l lst -in:file:fullatom -ignore_unrecognized_res -fa_max_dis 9.0 -database /mnt/home/dolgikhb/Rosetta/rosetta_bin_linux_2020.08.61146_bundle/main/database/ -ddg::harmonic_ca_tether 0.5 -score:weights pre_talaris_2013_standard -restore_pre_talaris_2013_behavior -ddg::constraint_weight 1.0 -ddg::out_pdb_prefix min_cst_0.5 -ddg::sc_min_only false -score:patch /mnt/home/dolgikhb/Rosetta/rosetta_bin_linux_2020.08.61146_bundle/main/database/scoring/weights/score12.wts_patch
Hello,
I have RESOLVED this issue! This post is for anyone who encounters a similar problem, and I will try to explain to the best of my knowledge.
The issue seemed to have been related to the convert_to_cst_file.sh command file and how the grep found the corrrect starting point from which to search for information in the lines of interest (containing generated constraint information). The original command was written as follows:
Where <c-alpha> should be the first word in the line, but was actually the second in the mincst.log file, thus resulting in an empty constraint file after running the command as <tcsh conver_to_cst_file.sh mincst.log > input.cst>
The resulting fix below allowed to extract the correct information for creating the constraint file by simply changing <^c-alpha> to the correct starting word and adjusting the numbering ($1, $2, $3, etc...) for the awk print portion:
I am unsure whether there are any changes between Rosetta versions in how the resulting minimize_with_cst mincst.log file looks (I am using 3.12), but I hope that if there are differences this post can sufficiently explain how to fix this problem, especially for naive programmers such as me!
I hope this helps! Good Luck!