BWA Sampe error - [bns_restore_core] fail to open file '1888]

[jomaco]

Hi

I ran bwa aln and it completed successfully. I have tried it numerous times and the same file sizes result so I assume this step is working correctly and the files are not truncated. However, when performing bwa sampe, I receive the following error:

------------------------------------------------------------
# LSBATCH: User input
source bwa-0.6.1; bwa sampe upregoakextractreal.fa LIB1167upregoakextract49.sai LIB1167upregoakextract49b.sai LIB1167_NoIndex_L005_R1.fastq LIB1167_NoIndex_L005_R2.fastq > L1167upregoakextract.sam
------------------------------------------------------------

Exited with exit code 134.

Resource usage summary:

    CPU time   :      0.19 sec.
    Max Memory :         2 MB
    Max Swap   :        31 MB

    Max Processes  :         1
    Max Threads    :         1

The output (if any) follows:

[bns_restore_core] fail to open file '1888]'. Abort!
/users/cd/jom/.lsbatch/1360003171.2694815: line 8: 28457 Aborted                 (core dumped) bwa sampe upregoakextractreal.fa LIB1167upregoakextract49.sai LIB1167upregoakextract49b.sai LIB1167_NoIndex_L005_R1.fastq LIB1167_NoIndex_L005_R2.fastq > L1167upregoakextract.sam
"saisam1.out" 40011L, 2158332C

I am not sure what file '1888]' refers to and I have tried numerous alignments and different versions of BWA to no avail. What might the problem be?

Thanks

Jom

[seb567]

Hi

I ran bwa aln and it completed successfully. I have tried it numerous times and the same file sizes result so I assume this step is working correctly and the files are not truncated. However, when performing bwa sampe, I receive the following error:

------------------------------------------------------------
# LSBATCH: User input
source bwa-0.5.10;bwa sampe upregoakextractreal.fa LIB1167upregoakextract49.sai LIB1167upregoakextract49.sai LIB1167_NoIndex_L005_R1.fastq LIB1167_NoIndex_L005_R2.fastq > L1167upregoakextract.sam
------------------------------------------------------------

Exited with exit code 134.

Resource usage summary:

    CPU time   :      0.19 sec.
    Max Memory :         2 MB
    Max Swap   :        31 MB

    Max Processes  :         1
    Max Threads    :         1

The output (if any) follows:

[bns_restore_core] fail to open file '1888]'. Abort!
/users/cd/jom/.lsbatch/1360003171.2694815: line 8: 28457 Aborted                 (core dumped) bwa sampe upregoakextractreal.fa LIB1167upregoakextract49.sai LIB1167upregoakextract49.sai LIB1167_NoIndex_L005_R1.fastq LIB1167_NoIndex_L005_R2.fastq > L1167upregoakextract.sam
"saisam1.out" 40011L, 2158332C

I am not sure what file '1888]' refers to and I have tried numerous alignments and different versions of BWA to no avail. What might the problem be?

Thanks

Jom

Can you provide the command line you used to invoke the program ?

[xied75]

Hi, Jom,

Did you notice that in your command your two sai files are the same name, but it should be one sai from read 1, another sai from read2.

This should have nothing to do with the '1888]' stuff though.

As a best practice, before you throw your command into any batch system, please test it does run correctly in terminal.

Best,

dong

[jomaco]

Hi

Thanks for your reply. I ran the following set of commands to get to this point:

source bwa-0.6.1;

bwa index -a is upregoakextractreal.fa

bsub -o saiout1 -n 8 -R "rusage[mem=40000] span[ptile=8]" 'source bwa-0.6.1; bwa aln -t 8 upregoakextractreal.fa LIB1167_NoIndex_L005_R1.fastq > LIB1167upregoakextract49.sai'    

bsub -o saiout1b -n 8 -R "rusage[mem=40000] span[ptile=8]" 'source bwa-0.6.1; bwa aln -t 8 upregoakextractreal.fa LIB1167_NoIndex_L005_R2.fastq > LIB1167upregoakextract49b.sai'   

bsub -o saisam1.out -R "rusage[mem=20000]" 'source bwa-0.6.1; bwa sampe upregoakextractreal.fa LIB1167upregoakextract49.sai LIB1167upregoakextract49b.sai LIB1167_NoIndex_L005_R1.fastq LIB1167_NoIndex_L005_R2.fastq > L1167upregoakextract.sam'

xied75: sorry I just noticed this, the correct command is now above. As I've tried running the command numerous times I originally ran it for files with different endings (i.e. not 49) and forgot to replace the 'b'. The same error still results with the correct command however.

Thanks

Jom

[jomaco]

I think I have finally solved the problem. The reference sequences I was aligning to are upregulated transcripts derived from transcriptome data. The output files from trinity had a format such as this (where some of the headers were very long):

>comp83983_c0_seq2 len=1889 path=[2318:0-132 2451:133-187 13407:188-216 2535:217-319 7725:320-374 7780:375-399 11674:400-402 2721:403-427 2746:428-441 2760:442-444 2763:445-479 13610:480-492 13623:493-497 13628:498-503 13634:504-504 10432:505-518 2837:519-521 2840:522-720 13735:721-739 7288:740-766 3097:767-795 13803:796-812 13818:813-831 13837:832-840 13846:841-851 13857:852-855 13861:856-860 13866:861-863 13869:864-867 13894:868-868 11863:869-900 3231:901-969 13934:970-970 13935:971-979 14604:980-1027 5920:1028-1038 3369:1039-1045 3376:1046-1084 14071:1085-1096 14052:1097-1109 14065:1110-1110 14066:1111-1115 3446:1116-1159 3490:1160-1191 3522:1192-1197 3528:1198-1344 5464:1345-1360 5480:1361-1373 14253:1374-1384 5699:1385-1396 3727:1397-1405 3736:1406-1451 14309:1452-1455 3786:1456-1483 14334:1484-1485 14336:1486-1493 14369:1494-1494 14345:1495-1506 14357:1507-1516 14367:1517-1560 14426:1561-1575 3906:1576-1579 3910:1580-1607 3938:1608-1615 3946:1616-1653 3984:1654-1740 4071:1741-1753 4084:1754-1779 4110:1780-1781 4112:1782-1805 14569:1806-1833 14661:1834-1864 14679:1865-1870 14685:1871-1884 14699:1885-1888]
GAAATTTCGCATGACTTAGTCAACTACGTAAACTGGGAACCATCACCCTTGATATATCTTAATAAACTTTAACATTGGAAACACAATACTTATCATCACGAAGTAGAATTTAAAATTTCCATAATTTTGGTCCTATCTCTTTTATGTCCTAAGCAATGAAGCTCTCACTTCTTTCCCCTTTTTCTCTAGGCTTACTCATAAAAATCTTATTCTTTTTCATCGCTGCTATAACACCACAAGCCGTACATCATGCTGATTACTCAAAGTGTCGAGAACAGTTCACGTGCGGATCAGTTAAGGATATTAAGTATCCTTTTTGGGGAAGTGATCGCCCTTCCTCTTGTGGCCTTGAAGGTTACAAAATCGAGTGCCAAGATAACGAGCCGATTTTCCATATGAATTGGCAACCATTTGTGGTCCGTAATATTGAAAGACCACCTAATTATACCATGGTAATTGCACGAGAAGATTTGTCTTACGATATTTGTCCCTCAAATTTTATTGATACCACCATCGATTTTAACCTTGTGAACTATAGCTCGACCGCCAAAAATCTCACCTTATTCTACTTATGTACCTCGGATGTTTCGGATAGTGCAGAACCAAACGAAGAGTTTAAATATTGCAGAGATAATCATAGTGTTTTTGGCATTTTAGATTTCCCGGAAAATAAAGTTTGGAGTACAGCGAGTCGAGAGTGTAAGAAGAAGATAAAAGTTCCGGTTATGCTAGAGAACAAGTTATCGTTAACAGAATTGATAAAGAACGGATTTGAAGTAGGTTACGAAACGCAGGCGAAAATGAAAAATTGCGAAACTT

I decided it must be the input files as I have now had successful results using exactly the same method for other reference sequences. The format of the input file as a whole looked fine but removing the sequence with the header which ended in '1888]' produced an error where bwa was looking for a different file instead so it must have been this which was causing the problem. Why bwa was trying to find this file, however, is beyond me. I went ahead and used sed to find and remove the path* part of the header for each sequence. The sampe stage of the alignment is now running and seems to be going fine. Before it crashed after just a few seconds - so fingers crossed!

Thanks for taking the time to look at my query

Jom

[atelford]

Hi there,

I know this thread was quite a while ago, but I am having the same problem, and although I have tried to remove the path* part of the header from the reference .fasta file (as per the last message), I can't find a way to do this. I have managed to remove 'Path=' but not the remaining '[5:7930 478:117890] etc. Sorry about this, I am a newbie and have spent quite a while trying to track this down but have hit a brick wall!

On a slightly separate note - does this problem not occur with everyone who is using assemblies from Trinity as a reference for bwa? I'm surprised I haven't found it referred to anywhere else!

Thanks very much

[GenoMax]

Hi there,

I know this thread was quite a while ago, but I am having the same problem, and although I have tried to remove the path* part of the header from the reference .fasta file (as per the last message), I can't find a way to do this. I have managed to remove 'Path=' but not the remaining '[5:7930 478:117890] etc. Sorry about this, I am a newbie and have spent quite a while trying to track this down but have hit a brick wall!

Thanks very much

If you are able to work with perl scripts then following will offer ways to shorten the perl headers: http://www.bioinformatics-made-simpl...ta-header.html

http://dawgpaws.sourceforge.net/man/fasta_shorten.html (package is at: http://sourceforge.net/projects/dawgpaws/)

I remember some awk/sed solutions being proposed (either here or on biostars) but I can't locate the relevant threads at the moment.

[atelford]

Thanks for that, I'll have a look and let you know how I get on.

I think the main problem I have been facing so far is that the part of the header I want to remove is variable - sometimes there are only a few numbers in the square parentheses, and sometimes there are loads - this is what I have had trouble trying to remove! Any suggestions would be gratefully received!

[GenoMax]

Thanks for that, I'll have a look and let you know how I get on.

I think the main problem I have been facing so far is that the part of the header I want to remove is variable - sometimes there are only a few numbers in the square parentheses, and sometimes there are loads - this is what I have had trouble trying to remove! Any suggestions would be gratefully received!

http://dawgpaws.sourceforge.net/man/fasta_shorten.html has the -l option that seems to allow you to specify the length to shorten to. Default is 20 but you can experiment with the length to keep everything before PATH (long as that part is unique).

[martin2]

Still happens with bwa 0.7.17-1188 when a sequence in FASTA file has a trailing space.

```
[bns_restore_core] Parse error reading somereference.fasta.amb
```