You are here

Rosetta_cm Segmentation Fault

3 posts / 0 new
Last post
Rosetta_cm Segmentation Fault
#1

Hello,

I wanted to make a separate post about another problem I'm having with comparative modeling. I am trying to use the M4 receptor as my target sequence and model it using the NOP receptor as my template structure. However, I am getting segmentation fault errors, which I have only previously gotten if my target FASTA sequence had line breaks in it. My threaded pdb file looks normal but for some reason I am getting the segmentation fault. The only thing I may be able to think of is that maybe the breaks in the loops in the threaded pdb are causing the fault, but am unsure. Here is my command and the output that follows:

 

[gszwbwsk@log002 M4R_NOP]$ /public/apps/rosetta/2017.29.59598/main/source/bin/rosetta_scripts.static.linuxgccrelease @rosetta_cm.options
core.init.score_function_corrections: [ WARNING ] Flag -restore_talaris_behavior is set but -weights are also specified.  Not changing input weights file!
core.init: Rosetta version unknown:89c6596b058f051ea67e56436a0486e432781ed7 2017-07-14 03:20:48 -0400 from /home/benchmark/release/rosetta/git/release/rosetta.binary.linux.release.git
core.init: command: /public/apps/rosetta/2017.29.59598/main/source/bin/rosetta_scripts.static.linuxgccrelease @rosetta_cm.options
core.init: 'RNG device' seed mode, using '/dev/urandom', seed=1323307466 seed_offset=0 real_seed=1323307466
core.init.random: RandomGenerator:init: Normal mode, seed=1323307466 RG_type=mt19937
core.init: Resolved executable path: /public/apps/rosetta/2017.29.59598/main/source/build/src/release/linux/3.10/64/x86/gcc/4.8/static/rosetta_scripts.static.linuxgccrelease
core.init: Looking for database based on location of executable: /public/apps/rosetta/2017.29.59598/main/database/
protocols.evaluation.ChiWellRmsdEvaluatorCreator: Evaluation Creator active ...
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch VirtualBB as requested
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch N_acetylated as requested
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch C_methylamidated as requested
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch ser_phosphorylated as requested
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch thr_phosphorylated as requested
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch ShoveBB as requested
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch LowerDNA as requested
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch UpperDNA as requested
core.chemical.GlobalResidueTypeSet: While generating GlobalResidueTypeSet centroid: Skipping patch VirtualDNAPhosphate as requested
core.chemical.GlobalResidueTypeSet: Finished initializing centroid residue type set.  Created 62 residue types
core.chemical.GlobalResidueTypeSet: Total time to initialize 0.04 seconds.
Segmentation fault (core dumped)
 

I realize that I am using an older version of Rosetta that we have on my school's computing cluster, but even the latest release on my local desktop gives me the same error. I have attached the cleaned pdb, post-threading pdb, and everything I'm using sequence and options-wise to this post. I know the causes of segmentation faults can be quite vague but I am at a loss here.

Post Situation: 
Thu, 2019-06-20 12:55
gszwabowski

Your XML file is specifying a disulfide file, but from your alignment and your threaded PDB, it looks like you don't have the cysteine partner that you would expect for a disulfide in ECL2 (or at least your alignment doesn't seem to indicate it.) -- Though it does look like there's a cysteine at 185 or so which might actually be aligned to the disulfied Cys in NOP.

I'd highly recommend double checking your alignment file, to make sure it's as good as it can be.  The one key factor for the quality of homology modeling results is the quality of the input alignment. It's well worth spending some time manually refining your algnment file, making sure your alignment is as high quality as possible, and incorporates all the information at your disposal regarding structural information, homologous positions, and family-specific correspondences. -- Note that the driver of the alignment file should be *structural* alignment and correspondence, not simply sequence alignment. (The best homology modeling alignment isn not necessarly the best BLOSUM62-scoring alignment.)

If you do believe that those cysteines aren't corresponding to a disulfide structure, then be sure to adjust the XML by removing the (unneeded) disulfide file.

I don't know if that's directly causing your segfault, but issues with disulfide specifications is a likely possibility.

Mon, 2019-06-24 13:28
rmoretti

Removal of the disulfide file specifications in the rosetta_cm.xml file still results in a segfault, unfortunately. Do you have any other suggestions?

Wed, 2019-06-26 10:25
gszwabowski