search

soedinglab / hh-suite

Remote protein homology detection suite.
https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-019-3019-7
GNU General Public License v3.0
529 stars 133 forks source link

Problem HHSuite - Segmentation fault #199

Closed dcerdanv closed 4 years ago

dcerdanv commented 4 years ago

:exclamation: Make to check out our User Guide.

Expected Behavior

Run hhblits and then addss to optain a .a3m file.

Current Behavior

When it is running hhblits it works properly until it is stopped due to a "Segmentation fault (core dumped)" always in the same point for every file. After that addss can be executed with no problem.

Steps to Reproduce (for bugs)

It is hard to reproduce the full context of the program, but the call to hhblits (and then addss.pl to add the secondary structure) is just: hhblits -i $dir/$chad.chd -d $DB_dir/$seq_DB/$seq_DB -oa3m $dir_output/$chad.a3m perl $directo/scripts/addss.pl $dir_output/$chad.a3m

I tried to change the -maxres, -maxmen and -cpu parameters as it is said in the wiki; and also I let to overcomit the memory (vm.overcommit_memory=1) but the problem persists.

HH-suite Output (for bugs)

The output of one call is:

Call to hhblits!!!

Segmentation fault (core dumped)

Call to addss!!! $ cp /localHDD/nuevo_fireDB/firedb_actualizacion/firestar/fdbrun/HH_DB_update/a3m_28Jan2020/5xg2A.a3m /tmp/qtl6K_lQfo/IqRjIRQrGM.in.a3m Filtering alignment to diversity 7 ... $ hhfilter -v 1 -neff 7 -i /tmp/qtl6K_lQfo/IqRjIRQrGM.in.a3m -o /tmp/qtl6K_lQfo/IqRjIRQrGM.in.a3m $ /usr/share/hhsuite/scripts/reformat.pl -v 1 -r -noss a3m psi /tmp/qtl6K_lQfo/IqRjIRQrGM.in.a3m /tmp/qtl6K_lQfo/IqRjIRQrGM.in.psi $ gunzip -c /localHDD/nuevo_fireDB/firedb_actualizacion/firestar/DB/GOLDEN_PDB/divided/xg/pdb5xg2.ent.gz > /tmp/1Niv5ezpx6 $ /usr/share/hhsuite/scripts/../../dssp-2.0.4/bin/dssp -i /tmp/1Niv5ezpx6 -o /tmp/qtl6K_lQfo/IqRjIRQrGM.dssp 2> /dev/null

Adding DSSP state sequence ... Predicting secondary structure with PSIPRED ... $ /usr/share/hhsuite/scripts/../../blast-2.2.26/bin/blastpgp -b 1 -j 1 -h 0.001 -d /usr/share/hhsuite/data/do_not_delete -i /tmp/qtl6K_lQfo/IqRjIRQrGM.sq -B /tmp/qtl6K_lQfo/IqRjIRQrGM.in.psi -C /tmp/qtl6K_lQfo/IqRjIRQrGM.chk 1> /tmp/qtl6K_lQfo/IqRjIRQrGM.blalog 2> /tmp/qtl6K_lQfo/IqRjIRQrGM.blalog $ echo IqRjIRQrGM.chk > /tmp/qtl6K_lQfo/IqRjIRQrGM.pn

$ echo IqRjIRQrGM.sq > /tmp/qtl6K_lQfo/IqRjIRQrGM.sn

$ /usr/share/hhsuite/scripts/../../blast-2.2.26/bin/makemat -P /tmp/qtl6K_lQfo/IqRjIRQrGM $ /usr/share/hhsuite/scripts/../../psipred-4.01/bin/psipred /tmp/qtl6K_lQfo/IqRjIRQrGM.mtx /usr/share/hhsuite/scripts/../../psipred-4.01/data/weights.dat /usr/share/hhsuite/scripts/../../psipred-4.01/data/weights.dat2 /usr/share/hhsuite/scripts/../../psipred-4.01/data/weights.dat3 > /tmp/qtl6K_lQfo/IqRjIRQrGM.ss $ /usr/share/hhsuite/scripts/../../psipred-4.01/bin/psipass2 /usr/share/hhsuite/scripts/../../psipred-4.01/data/weights_p2.dat 1 0.98 1.09 /tmp/qtl6K_lQfo/IqRjIRQrGM.ss2 /tmp/qtl6K_lQfo/IqRjIRQrGM.ss > /tmp/qtl6K_lQfo/IqRjIRQrGM.horiz done

Notice than "Call to hhblits" or "Call to addss" are just two sentences that I printed to know what was failing. At the end I get the .a3m file, but probably incomplete because of the fault.

Your Environment

I'm runing this in Ubuntu 18.04. In a 16 cores machine, each of them is a Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz, and a RAM of 94GB. I don't use any special enviroment, I tried with conda but I get more issues so I install everything directly in my machine. I'm using:

Thank you very much in advance.

milot-mirdita commented 4 years ago

Could you upload this 5xg2A.chd file? I can try to reproduce the issue with it.

dcerdanv commented 4 years ago

Do not worry milot-mirdita, it seens to be a problem of permissions. When I runned it as sudo it gave memory problems, and when not it gave other differents. Finally I could install it in conda and run it from there, without "sudo", and it is working now.

Thank you very much.