Entering edit mode
7.1 years ago
anu014
▴
190
Greetings of the day Biostars,
Recently, I am stuck with this error :
Segmentation fault (core dumped) (ERR): bowtie2-align exited with value 139.
I've checked all over the internet, but no solution is working for me. I checked my FASTQ files too with 'validateFastq' : no errors. I tried running bowtie2 on 2 different servers with high cores n memory but still same error.
Please help me out.
Thank you in advance :)
What version?
$bowtie2 --version
/usr/bin/bowtie2-align-s version 2.2.6 64-bit Built on lgw01-12 Mon Dec 28 11:09:46 UTC 2015 Compiler: gcc version 5.3.1 20151219 (Ubuntu 5.3.1-4ubuntu1) Options: -O3 -m64 -msse2 -funroll-loops -g3 -Wl,-Bsymbolic-functions -Wl,-z,relro -DPOPCNT_CAPABILITY -DWITH_TBB Sizeof {int, long, long long, void*, size_t, off_t}: {4, 8, 8, 8, 8, 8}
Try the most recent version. If that doesn't work, try only aligning the first million reads, then the second million, and so on to see if this is due to something specific read/read-pair in the input.
bowtie2 v.2.3.1 is out. Would not hurt to try to upgrade and see if it helps.
I tried with 2.3.1v but it also showed same error.. Below is the command I ran:
nohup /home/anupriya/bowtie2-2.3.1/bowtie2 -p 30 --phred33 --local --sensitive -x /home/anupriya/bowtie2_index/ref_index -1 L002_R1_001_val_1.fq -2 L002_R2_001_val_2.fq -S L002.sam > nohup_trial_002.txt
Do you actually have 30 threads available on this server? Can you dial that back to say 8 and see what happens? What OS are you running this on and did you compile bowtie2/downloaded the pre-compiled binary?
$uname -a Linux carpo 4.4.0-64-generic #85-Ubuntu SMP Mon Feb 20 11:50:30 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux .
$lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 16.04.2 LTS
Release: 16.04
Codename: xenial
I downloaded bowtie2-2.3.1-linux-x86_64.zip file from http://bowtie-bio.sourceforge.net/bowtie2/manual.shtml#building-from-source & used the command 'make' as written on the website.
On the plus side, you downloaded the precompiled binary rather than the source code. On the minus side, you apparently thought that needed to be compiled.
Anyway, you'll need to do what I suggested in my comment above.
My paired-end files was having 31084508 line each. So I divided files into 7 smaller pairs, with 6 pairs containing 5000000 & last pair containing 1084508 lines each. The 4th segmented files threw same error "Segmentation fault (core dumped) (ERR): bowtie2-align exited with value 139". What to do now? @Devon Ryan
If it was only that one pair that threw the error, then subdivide that file again. You can do that until you have a very small (say 100 read pairs) file that reproduces the segfault. If visual inspection suggests that the files are then fine (this will likely be the case given what you've written), then you already have a perfect example test-case for a bug report. The bowtie2 authors aren't terribly responsive in my experience, but if you report a bug and give them a convenient example dataset then they'll be quicker to fix things.
Additionally, one can then merge all of the BAM files generated together and then continue on with the analysis with minimal loss of reads.
Hi Devon,
I did as you suggested. Got small paired file for 25 reads still unable to find the error.. I guess I have to send the bug report to Bowtie authors. I did one more thing. I ran complete data with BWA-MEM (default parameters) & everything ran succesfully! Reason unknown..
Weird, this is totally a bowtie2 bug then. Glad you at least got your alignments!
Running on 8 cores didn't work..
That does not make sense and should not happen.
Edit: Unless you are spreading the threads from your job across multiple physical nodes and disk I/O limitations is causing some strange issues. You should keep threads from aligners restricted to single node (since most of them are not truly parallel).