You are here

demos/public/ideal_proteins Segmentation fault

6 posts / 0 new
Last post
demos/public/ideal_proteins Segmentation fault
#1

Dear All,

I am using rosetta_bin_linux_2017.52.59948_bundle to run demos/public/ideal_proteins. The running command is 

"rosetta_scripts.static.linuxgccrelease -s input.pdb -parser:protocol input.xml -nstruct 10 -restore_talaris_behavior".

Since I got 

"

score_typeextract failed: fa_plane

 

ERROR: bad line in file myweight.wts:fa_plane 0

ERROR:: Exit from: src/core/scoring/ScoreFunction.cc line: 453

" I comment out fa_plane in myweight.wts for running

Here are things I observed:

verion 2017.52, Linux (centos 6/7 both static and dynamic) got "Segmentation fault (core dumped)".

verion 2017.52, macos runs perfectly.

released 3.8, Linux runs perfectly.

My question is why Linux verion 2017.52 cannot run?

Many thanks

Category: 
Post Situation: 
Wed, 2018-03-07 00:15
coomteng@gmail.com

The `fa_plane` term wasn't actually doing anything, and wasn't actually used, so we removed it. Unfortunately, it looks like we forgot to remove it from the myweight.wts file in the ideal_proteins demo.  -- But that's okay, as it's set to a weight of zero (i.e. turned off) anyway.

You should be able to just delete that line from the weights file to solve that particular issue with 2017.52.

(I'm not sure why 2017.52 on Linux is giving you a segmentation fault -- I'm not sure if it's related to this or not.)

Wed, 2018-03-07 09:04
rmoretti

The fa_plane fix is in 3.9 but not 3.8; it surely made it in by 2017.52.

Given that you hae two linux versions you have floating around it could be a database error, if you are using the 2017.52 database with 3.8 or vice versa.

Wed, 2018-03-07 09:13
smlewis

The fa_plane fix is in the code and the main database for 2017.52, but the particular demo they're using has its own weights file which still had the fa_plane term present at zero weight. (This is now fixed in the developer's copy, but is not yet in any released version.)

Thu, 2018-03-08 07:46
rmoretti

Thank you for your reply.

The installation was new, there should not be mix-up of database usage.

 

I compare the ourput of Linux and MAC verion of 2017.52

Linux:
protocols.rosetta_scripts.ParsedProtocol: =======================BEGIN MOVER LoopOver - lover1=======================
protocols.protein_interface_design.movers.LoopOver: Loop iteration 0
protocols.fldsgn.BluePrintBDR: SegmentRebuild left: 1, right: 2, ss: LEEEEEEELLLHHHHHHHHHHHHHHHHHHLLEEEEEEELLEEEEEEEL, aa:VVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVV
core.conformation.Residue: [ WARNING ] Residue connection id changed when creating a new residue at seqpos 1
core.conformation.Residue: [ WARNING ] ResConnID info stored on the connected residue (residue 2) is now out of date!
core.conformation.Residue: [ WARNING ] Connection atom name (in src):  C
protocols.forge.build.BuildManager: FOLD_TREE  EDGE 1 1 -1  EDGE 1 48 -1
protocols.fldsgn.BluePrintBDR: ResiduePairConstraints: total: 0   plotting active...
protocols.fldsgn.BluePrintBDR: IntraResidueConstraints: total: 0 showing active...
protocols.fldsgn.BluePrintBDR: NonResiduePairConstraints: total: 0 showing active...
core.conformation.Residue: [ WARNING ] Residue connection id changed when creating a new residue at seqpos 1
core.conformation.Residue: [ WARNING ] ResConnID info stored on the connected residue (residue 2) is now out of date!
core.conformation.Residue: [ WARNING ] Connection atom name (in src):  C
protocols.forge.build.BuildManager: FOLD_TREE  EDGE 1 1 -1  EDGE 1 48 -1
protocols.forge.components.VarLengthBuild: Abego size=48 and pose size=48
Segmentation fault

 

MAC:

protocols.rosetta_scripts.ParsedProtocol: =======================BEGIN MOVER LoopOver - lover1=======================
protocols.protein_interface_design.movers.LoopOver: Loop iteration 0
protocols.fldsgn.BluePrintBDR: SegmentRebuild left: 1, right: 2, ss: LEEEEEEELLLHHHHHHHHHHHHHHHHHHLLEEEEEEELLEEEEEEEL, aa:VVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVV
core.conformation.Residue: [ WARNING ] Residue connection id changed when creating a new residue at seqpos 1
core.conformation.Residue: [ WARNING ] ResConnID info stored on the connected residue (residue 2) is now out of date!
core.conformation.Residue: [ WARNING ] Connection atom name (in src):  C  
protocols.forge.build.BuildManager: FOLD_TREE  EDGE 1 1 -1  EDGE 1 48 -1
protocols.fldsgn.BluePrintBDR: ResiduePairConstraints: total: 0   plotting active...
protocols.fldsgn.BluePrintBDR: IntraResidueConstraints: total: 0 showing active...
protocols.fldsgn.BluePrintBDR: NonResiduePairConstraints: total: 0 showing active...
core.conformation.Residue: [ WARNING ] Residue connection id changed when creating a new residue at seqpos 1
core.conformation.Residue: [ WARNING ] ResConnID info stored on the connected residue (residue 2) is now out of date!
core.conformation.Residue: [ WARNING ] Connection atom name (in src):  C  
protocols.forge.build.BuildManager: FOLD_TREE  EDGE 1 1 -1  EDGE 1 48 -1
protocols.forge.components.VarLengthBuild: Abego size=48 and pose size=48
core.chemical.GlobalResidueTypeSet: Finished initializing centroid residue type set.  Created 62 residue types
core.chemical.GlobalResidueTypeSet: Total time to initialize 0.04676 seconds.
protocols.forge.components.VarLengthBuild: VLB count_cutpoints 0 interval.left 1 interval.right 48
protocols.forge.components.VarLengthBuild: picking 200 9-mers for position 1
protocols.forge.components.VarLengthBuild: complete_ss length: 48 complete_abego size: 48
core.fragment.picking_old.vall.vall_io: Reading Vall library from /Users/zhouti8/software/rosetta/rosetta_bin_mac_2017.52.59948_bundle/main/database//sampling/filtered.vall.dat.2006-05-05 ...
core.fragment.picking_old.vall.vall_io:    100000
core.fragment.picking_old.vall.vall_io:    200000
core.fragment.picking_old.vall.vall_io:    300000

It looks like Linux version stuck at "core.chemical.GlobalResidueTypeSet: Finished initializing centroid residue type set.  Created 62 residue types
core.chemical.GlobalResidueTypeSet: Total time to initialize 0.04676 seconds."

Wed, 2018-03-07 13:46
coomteng@gmail.com

if I change a line input.xml <BluePrintBDR name="bdr1" scorefxn="SFXN1" use_abego_bias="1" blueprint="./input1.bbskel" />

into <BluePrintBDR name="bdr1" scorefxn="SFXN1" use_abego_bias="0" blueprint="./input1.bbskel" />, then it can run.

The content of input1.bbskel is
1 V L R
2 V E R
0 V E R
0 V E R
0 V E R
0 V E R
0 V E R
0 V E R
0 V L R
0 V L R

...

which does not have "abego_bias".

Wed, 2018-03-07 14:43
coomteng@gmail.com