metaspades error code -11

[metalichen]

Hi guys,

I've run into an issue running SPAdes in metagenomic mode. Filtering singleton k-mers of the forward reads takes only a few minutes, but then SPAdes gets stuck on the reverse reads file for a few hours and finally finishes with error.

Here is the command I used:

/core/software/conda/bin/metaspades.py -k 21,33,55,77,99 --pe1-1 ./trimmed_reads/VT31_S30_L006_R1_001_pairedPE.fq.gz --pe1-2 ./trimmed_reads/VT31_S30_L006_R2_001_pairedPE
.fq.gz -t 12  -o spades_out/VT31

And this is what I got:

 0:02:59.814    48M / 240M  INFO   K-mer Counting           (kmer_data.cpp             : 311)   Filtering singleton k-mers
38 8 0
nslots: 1073741824
bits per slot: 8 range: 0000004000000000
  0:02:59.814     1G / 1G    INFO   K-mer Counting           (kmer_data.cpp             : 317)   Processing /scratch/1/gulnara/2018libraries/wo-00000469/trimmed_reads/VT31_S30_L006_R1_001_pairedPE.fq.gz
  0:07:46.942     1G / 1G    INFO   K-mer Counting           (kmer_data.cpp             : 326)   Processed 21126117 reads
  0:07:46.942     1G / 1G    INFO   K-mer Counting           (kmer_data.cpp             : 317)   Processing /scratch/1/gulnara/2018libraries/wo-00000469/trimmed_reads/VT31_S30_L006_R2_001_pairedPE.fq.gz
The program was terminated by segmentation fault
 === Stack Trace ===

== Error ==  system call for: "['/core/software/conda/share/spades-3.12.0-1/bin/spades-hammer', ' /scratch/1/gulnara/2018libraries/wo-00000469/spades_out/VT31/corrected/configs/config.info']" finished abnormally, err code: -11

There is one weird thing about this issue. I get it when I'm trying to assembly only one library. Other libraries coming from the same sequencing lane are just fine, although all of them were treated absolutely the same way and the number of reads doesn't vary that much. By the same way I mean quality trimming by Trimommatic 0.33. I thought that the problem could be in that library itself, but when I tried to assemble it as it was with no trimming, I didn't get any error.

Here is the link to the data and spades.log file. https://drive.google.com/open?id=1mJiIYjxwapg_GA669OCz2RnVALb3qlDZ

Any ideas what could be wrong?

Thanks!

[davidbio]

Hi,

I get the same error here with a different setup:

• --meta mode
• 1M reads paired-end
• parameters are in the log
• spades.log
• data 136 MB
• Please note that the time consumption is ridiculously high as well (12+ hours).

[asl]

@davidbio Please do not hijack other issues even if there are some similar words, the sources of problems might be completely different :) Please check whether the latest SPAdes 3.13 fixes the problem for you.

[davidbio]

@asl I thought opening another issue with the exact same error message would be counterproductive. I will open another issue then, I didn't mean to be rude.

[asl]

@davidbio The problem is: it was our fault to show these "error codes". In fact they do mean nothing, it's just the return code provided by the system when the program is terminated. Sometimes it's meaningful (e.g. -9 means that the job was killed via SIGKILL signal) but in majority of cases is misleading. Only log contains relevant information and the real issue might be completely different provided the same "error" code :)

So, please do try and reproduce issue with the latest SPAdes 3.13. If there is still an error – please open separate issue.

Thanks for understanding! :)

[kdbchau]

Was there a solution for this? I have the same issue but with Spades 3.15.1 verison.

[dsfchina]

Using spades 4.0.0 on my computer solves this problem: conda create spades conda activate spades conda install libgcc-ng=12.3.0 -c conda-forge conda install spades=4.0.0 -c bioconda