Error when running FRAGMENT generation

[Anonymous]

Dear all,

I try to follow protocol and want to have fragment picker components for running Flexpepdock ab-initio protocol. But program return the ERROR. Can someone give me solution.

Thank alot for that.

this is my flag:
-s SPI.pdb
-database /home/tuongvy/SW/rosetta3.4/rosetta_database

-vall vall.dat.apr24.combo.v2.mar10.bfactor.v2

-frags:bounded_protocol

-frags:n_frags 200
-frags:n_candidates 200

-frags:frag_sizes 3
-frags:scoring:config quota_scores.wghts

-out:file:frag_prefix
-frags:describe_fragments

and
quoto_scores.wghts file is

SecondarySimilarity 350 1.0 – psipred
SecondarySimilarity 300 1.0 – sam
SecondarySimilarity 250 1.0 – porter
RamaScore 150 2.0 – psipred
RamaScore 150 2.0 – porter
RamaScore 150 2.0 – sam
ProfileScoreL1 200 2.0 –
PhiPsiSquareWell 100 0.0 –
FragmentCrmsd 30 0.0 –

ERROR AFTER RUNNING:

[node08:00924] *** Process received signal ***
[node08:00924] Signal: Segmentation fault (11)
[node08:00924] Signal code: Address not mapped (1)
[node08:00924] Failing at address: (nil)
[node08:00924] [ 0] /lib64/libpthread.so.0 [0x3a77a0ebe0]
[node08:00924] [ 1] /home/tuongvy/SW/rosetta3.4/rosetta_source/build/src/release/linux/2.6/64/x86/gcc/4.1/mpi/libprotocols_a.2.so(_ZN9protocols11frag_picker6scores23MakeSecondarySimilarity4makeEmdbN7utility7pointer10owning_ptrINS0_14FragmentPickerEEESs+0x3c) [0x2b1f50803e3c]
[node08:00924] [ 2] /home/tuongvy/SW/rosetta3.4/rosetta_source/build/src/release/linux/2.6/64/x86/gcc/4.1/mpi/libprotocols_a.2.so(_ZN9protocols11frag_picker6scores20FragmentScoreManager21create_scoring_methodERKSsmddbN7utility7pointer10owning_ptrINS0_14FragmentPickerEEESs+0x24e) [0x2b1f507fa21e]
[node08:00924] [ 3] /home/tuongvy/SW/rosetta3.4/rosetta_source/build/src/release/linux/2.6/64/x86/gcc/4.1/mpi/libprotocols_a.2.so(_ZN9protocols11frag_picker6scores20FragmentScoreManager13create_scoresERKSsN7utility7pointer10owning_ptrINS0_14FragmentPickerEEE+0x56f) [0x2b1f507fb3df]
[node08:00924] [ 4] /home/tuongvy/SW/rosetta3.4/rosetta_source/build/src/release/linux/2.6/64/x86/gcc/4.1/mpi/libprotocols_a.2.so(_ZN9protocols11frag_picker14FragmentPicker18parse_command_lineEv+0x1a54) [0x2b1f50903914]
[node08:00924] [ 5] /home/tuongvy/SW/rosetta3.4/rosetta_source/bin/fragment_picker.mpi.linuxgccrelease(main+0x145) [0x406eb5]
[node08:00924] [ 6] /lib64/libc.so.6(__libc_start_main+0xf4) [0x3a76a1d994]
[node08:00924] [ 7] /home/tuongvy/SW/rosetta3.4/rosetta_source/bin/fragment_picker.mpi.linuxgccrelease(__gxx_personality_v0+0x191) [0x406969]
[node08:00924] *** End of error message ***
mpiexec noticed that process rank 7 with PID 924 on node node08.mark.pknu.ac.kr exited on signal 11 (Segmentation fault).

---

[Anonymous]

Fragment picking needs a sequence as input, not a structure. One of “-in::file::checkpoint”, “-in::file::binary_chk”, “-in::file::pssm” or “-in::file::fasta” should be provided.

If you don’t know what to provide the other options, I’d recommend sticking with the fasta sequence option. The script “main/source/src/python/apps/public/pdb2fasta.py” is able to extract a fasta sequence from a PDB file. (You will want to examine and possibly edit the output before using, though.)

[Anonymous]

Hi phanvy,
Please download the FlexPepDock tutorial available here: http://bit.ly/NfW6tq. You can set up fragment creation using the instructions given. You need to run PSIpred and Blast locally. From our analysis, we have see that PSIpred work well enough for small peptides, though it’s not optimized for small length. You need to edit make_fragments.pl file to include the paths of the binaries and databases. We only use PSIpred predicted secondary structure and blast output to get the fragments. Then fragment_picker is used to get the fragments. All are set up in prep_abinitio.sh script. Please go though te README file. You can have a look at the scoring method used to get the fragments. Please let us know if you have any queries.
best of luck!

[Anonymous]

Dear rmoretti
thank a lot for your susggestion.

I have the peptide only 7 amino acid, the sequence is so short and PsiPred webserve can not run (the error “the sequence is too short). How can I get secondary structure prediction of this peptide?

when I use http://distill.ucd.ie/porter/ to get secondary structure predition the result return like this :
TMEPGKP
CCCCCCC
But, the program can not rescogine this file.

I want to use BestFragmentProtocol
# Input databases
-database ../database
-in::file::vall ../vall-database/vall.apr24.2008.extended.gz

# Query-related input files
#-in::file::checkpoint input_files/2jsvX.checkpoint
-in::file::fasta input_files/2jsvX.fasta
-in::file: input_files/2jsvX.pdb
-frags::ss_pred input_files/2jsvX.psipred.ss2 predA

# Weights file
-frags::scoring::config input_files/simple.wghts

# What should we do?
-frags::bounded_protocol

# three-mers only, please
-frags::frag_sizes 3
-frags::n_candidates 200
-frags::n_frags 200

# Output
-out::file::frag_prefix output_files/frags

this protocol require psipred.ss2 file?

And another question :
With only .fasta file, Can I run fragment Picker?

Thank a lot.

[Anonymous]

Dear nawad
the website, somtimes, have the problem such the updating can not going on?
Could you please post again your reply

Best regards,
Phan Vy

[Anonymous]

Dear Naswad,
I try to follow the instructions, and try to run fragment_picker locally. Some the paths of binaries and databases can not edit because I can not find (such as : robetta/databases/local_db/nr/nr_pfilt, robetta/src/shareware/sam.predict2nd …).
And another question: when I used different pdb files for reference, I got the same result fragment. it means the pdb file fragment is just for caculting RMSD and it was not effect the picking and collecting processing, right?

[Anonymous]

Frgament picking relies on several pieces of external software, including the BLAST nr database, and things like sam, jufo, etc. You’ll need to install them locally, and then change the directory locations in the scripts to point to the locations on your machine.

Fragment picking is also primarily sequence based. So the different input structures will have little to no effect on the results of your fragment picking.

[Author]

Posts