[maker-devel] Strange error at blastn step

Fri Oct 26 12:19:06 MDT 2012

Hi Daniel,

What version or revision of MAKER are you running.  This sounds like something we were seeing here last night.  We traced it as far as what appeared to be soft links in /tmp being set incorrectly.  The FastaDB objects had pointers to fasta files in the void for the correct fasta file, but their dir attribute pointed to a /tmp directory in where there were soft links to another (incorrect)  fasta file with it's index.  It would look appeared that it was looking in the /tmp index (which pointed to an incorrect fasta file) and when it failed to find what it was looking for it would re-index the fasta file in the void (the correct one) and then look again in the index in tmp.  Don't know if that helps, but the errors look similar and that's as far as I got with our error here…  This was on one of Mike's annotation projects, so I don't know for sure what revision he was running, but I think it was the latest.

B

On Oct 26, 2012, at 11:52 AM, Daniel Standage wrote:

> I have since installed Maker on a different machine and tried it out. The test run completed successfully, but as I commenced with the full genome annotation, I have noticed the following error popping up frequently.
> 
> formating database...
> #--------- command -------------#
> Widget::formater:
> /N/u/dstandag/Mason/local/bin/makeblastdb -dbtype prot -in /tmp/maker_1YQF9o/Amel3%2E2_Dmel5%2E47%2Efaa.mpi.10.8
> #-------------------------------#
> running  blast search.
> #--------- command -------------#
> Widget::blastx:
> /N/u/dstandag/Mason/local/bin/blastx -db /tmp/maker_1YQF9o/Amel3%2E2_Dmel5%2E47%2Efaa.mpi.10.8 -query /tmp/maker_1YQF9o/rank0/scaffold_0.0 -num_alignments 100000 -num_descriptions 100000 -evalue 1e-06 -dbsize 300 -searchsp 500000000 -num_threads 16 -seg yes -soft_masking true -lcase_masking -show_gis -out /N/dc/scratch/dstandag/PdomGenomic/Annotation/output/maker.pdom.1.mason.maker.output/maker.pdom.1.mason_datastore/scaffold_0/theVoid.scaffold_0/scaffold_0.0.Amel3%2E2_Dmel5%2E47%2Efaa.blastx.temp_dir/Amel3%2E2_Dmel5%2E47%2Efaa.mpi.10.8.blastx
> #-------------------------------#
> deleted:-10 hits
> formating database...
> #--------- command -------------#
> Widget::formater:
> /N/u/dstandag/Mason/local/bin/makeblastdb -dbtype prot -in /tmp/maker_1YQF9o/Amel3%2E2_Dmel5%2E47%2Efaa.mpi.10.9
> #-------------------------------#
> running  blast search.
> #--------- command -------------#
> Widget::blastx:
> /N/u/dstandag/Mason/local/bin/blastx -db /tmp/maker_1YQF9o/Amel3%2E2_Dmel5%2E47%2Efaa.mpi.10.9 -query /tmp/maker_1YQF9o/rank0/scaffold_0.0 -num_alignments 100000 -num_descriptions 100000 -evalue 1e-06 -dbsize 300 -searchsp 500000000 -num_threads 16 -seg yes -soft_masking true -lcase_masking -show_gis -out /N/dc/scratch/dstandag/PdomGenomic/Annotation/output/maker.pdom.1.mason.maker.output/maker.pdom.1.mason_datastore/scaffold_0/theVoid.scaffold_0/scaffold_0.0.Amel3%2E2_Dmel5%2E47%2Efaa.blastx.temp_dir/Amel3%2E2_Dmel5%2E47%2Efaa.mpi.10.9.blastx
> #-------------------------------#
> deleted:-6 hits
> WARNING: Cannot find> comp59088_c1_seq7, trying to re-index the fasta.
> stop here:comp59088_c1_seq7
> ERROR: Fasta index error
> 
> FATAL ERROR
> ERROR: Failed while polishig ESTs!!
> 
> ERROR: Chunk failed at level 14
> !!
> FAILED CONTIG:scaffold_0
> 
> 
> 
> 
> --Next Contig--
> 
> #---------------------------------------------------------------------
> Now starting the contig!!
> SeqID: scaffold_1
> Length: 5805686
> #---------------------------------------------------------------------
> 
> My first thought based on the message is that blastdbcmd could not find the sequence in the database. I verified this was the case--I could not extract sequence comp59088_c1_seq7 from the database Maker had created under /tmp. However, after removing the index files and re-running makeblastdb with the -parse_seqids option set, blastdbcmd successfully extracted the sequence.
> 
> I was initially happy with this finding, but upon closer inspection it looks like Maker does not use blastdbcmd to extract sequences, but rather its own internal code. Therefore I'm still not sure where the problem is and how I might fix it. Any insights?
> 
> Thanks!
> 
> --
> Daniel S. Standage
> Ph.D. Candidate
> Bioinformatics and Computational Biology Program
> Department of Genetics, Development, and Cell Biology
> Iowa State University
> 
> 
> 
> On Thu, Oct 25, 2012 at 3:30 PM, Daniel Standage <daniel.standage at gmail.com> wrote:
> Greetings!
> 
> I am doing a test run of my Maker setup on a new machine, annotating a pretty short contig (about 3kb). However, there seems to be a hiccup during the blastn stage. This is the terminal message.
> 
> #--------- command -------------#
> Widget::blastn:
> /share/home/01854/standage/local/bin/blastn -db /tmp/2881473.1.development/maker_i6ZSJi/Pdom%2ETrinity%2ETrimmomatic%2Efasta.mpi.10.7 -query /tmp/2881473.1.development/maker_i6ZSJi/rank0/scaffold_866.0 -num_alignments 100000 -num_descriptions 100000 -evalue 1e-10 -gapextend 3 -word_size 15 -reward 1 -penalty -3 -gapopen 3 -dbsize 1000 -searchsp 500000000 -num_threads 16 -lcase_masking -dust yes -soft_masking true -show_gis -out /scratch/01854/standage/PdomGenomic/Annotation/scripts/maker.bogus.maker.output/maker.bogus_datastore/scaffold_866/theVoid.scaffold_866/scaffold_866.0.Pdom%2ETrinity%2ETrimmomatic%2Efasta.blastn.temp_dir/Pdom%2ETrinity%2ETrimmomatic%2Efasta.mpi.10.7.blastn
> #-------------------------------#
> deleted:0 hits
> ERROR:  Could not obtain lock to format database
> 
> 
> FATAL ERROR
> ERROR: Failed while doing blastn of ESTs!!
> 
> ERROR: Chunk failed at level 8
> !!
> FAILED CONTIG:scaffold_866
> 
> Several blastn steps appeared to have completed successfully to this one failing. Any ideas what could be causing this?
> 
> Thanks!
> 
> --
> Daniel S. Standage
> Ph.D. Candidate
> Bioinformatics and Computational Biology Program
> Department of Genetics, Development, and Cell Biology
> Iowa State University
> 
> 
> _______________________________________________
> maker-devel mailing list
> maker-devel at box290.bluehost.com
> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org

Barry Moore
Research Scientist
Dept. of Human Genetics
University of Utah
Salt Lake City, UT 84112
--------------------------------------------
(801) 585-3543

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://yandell-lab.org/pipermail/maker-devel_yandell-lab.org/attachments/20121026/ab5ff667/attachment-0003.html>