You are here

loopmodel issue

2 posts / 0 new
Last post
loopmodel issue
#1

I am trying to use loopmodel to fold a linear peptide of 216 standard aa (obtained from the only existing data: fasta file, by using either the BuildPeptide or FragmentPicker executables of rosetta3.3).

I first tried "loopmodel.mpi.linuxgccrelease" (in my Open MPI compilation :scons bin mode=release cxx=gcc cxx_ver=4.6" ALL executables are both ".mpi." and non-mpi, which is strange as not all are parallelized) according to the command line:

francesco@gig64:~/tmp$ /usr/local/rosetta_source/bin/loopmodel.mpi.linuxgccrelease -database /usr/local/rosetta_database -j6 mpi_work_partition_job_distributor -loops:remodel perturb_kic -loops:refine refine_kic -loops:input_pdb my.pdb -loops:loop_file modeling.loops -loops:outer_cycles 1 -loops:refine_init_temp 1.2 -loops:refine_final_temp 1.2 -loops:vicinity_sampling -loops:vicinity_degree 3 -loops:optimize_only_kic_region_sidechains_after_move -ex1 -ex2 -nstruct 100

then, the same for the non-mpi executable, while also removing "-j6 work.....distributor"

In BOTH CASES, no atom was recognized and MPI error came out:

.................................
................................

core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom CD (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom CG (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom NE2 (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom OE1 (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom OXT (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom 1HB (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom 1HE2 (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom 1HG (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom 2HB (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom 2HE2 (trying to set temp)
core.io.pdb.file_data: (0) [ WARNING ] can't find atom for res 216 atom 2HG (trying to set temp)
core.pose.util: (0) Cannot open psipred_ss2 file tt
protocols.loops.loops_main: (0) can not open DSSP file tt
protocols::loopbuild: (0) Annotated sequence of start_pose: M[MET_p:NtermProteinFull]KITEFMKNKKGAS........................................TEKVVSLQ[GLN_p:CtermProteinFull]

ERROR: [ERROR] Error opening RBSeg file 'modeling.loops'
ERROR:: Exit from: src/protocols/loops/Loops.cc line: 482
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD
with errorcode 911.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
francesco@gig64:~/tmp$

Thanks a lot for advice
chiendarret

Post Situation: 
Fri, 2012-01-27 00:51
chiendarret

This isn't an MPI error, this is clearly a rosetta error: "ERROR: [ERROR] Error opening RBSeg file 'modeling.loops'
ERROR:: Exit from: src/protocols/loops/Loops.cc line: 482"

So, something is wrong with your modeling.loops file - my first guess is that it's not present. My second guess is that it's not defining a loop properly?

As far as mpi executables - if mpi is in the _path_, like build/src/release/linux/2.6/x86/gcc/mpi/loopmodel.linuxgccrelease, then it's mpi, whether it has mpi in the filename or not. We fixed the double-executables issue.

Finally, you should be using ab initio to fold fasta sequences into proteins, not loop modeling. Loop modeling is for remodeling <=12 residue stretches with no alpha or beta structure, not for whole proteins at once.

Also, the -j flag has no effect on Rosetta - that's for compiling. Also also, for MPI, you need to use the mpirun supplied with your MPI installation - something like "mpirun -np6 rosetta"

Fri, 2012-01-27 05:42
smlewis