Segmentation fault (core dumped), err code: 255

[githubgig]

Hello,

I have two questions:

1. I got "Segmentation fault (core dumped)" message at bwa mem alignment step (the last few lines copied below my message and attached are params.txt and spades.log). Is this a memory issue or any issue due to read files being .fasta files? The genome I'm trying to assemble is 1.6 Gbp.

2. From the previous steps before the error/crash occurred, I have the assembled_contigs.fasta and assembled_scaffolds.fasta that are in the ../Output/misc folder. Do the steps starting from bwa mem (polishing etc) improve the assembly statistics (N50 etc.) any further or do the files assembled_contigs.fasta and assembled_scaffolds.fasta represent the last stage of 'assembly'? The reason I ask is that both Illumina and Pacbio that I input into SPADES were already error corrected. So is it necessary for me to continue to the final step or can I just use the assembled_scaffolds.fasta file as a substitute for scaffolds.fasta file?

Thanks.

Segmentation fault (core dumped) 0:08:31.463 628M / 632M INFO DatasetProcessor (dataset_processor.cpp : 152) bwa failed, skipping sublib 0:08:31.463 628M / 632M ERROR DatasetProcessor (dataset_processor.cpp : 215) Failed to align paired reads /home/user/Genome/Assemblies/SPADES/Lane12_R1_Cutadapt_QuorumCorr_MinLen100_BBNorm.fasta and /home/user/Genome/Assemblies/SPADES/Lane12_R2_Cutadapt_QuorumCorr_MinLen100_BBNorm.fasta

== Error == system call for: "['/home/user/bin/SPAdes-3.13.0-Linux/bin/spades-corrector-core', '/home/user/Genome/Assemblies/SPADES/Output/mismatch_corrector/contigs/configs/corrector.info', '/home/user/Genome/Assemblies/SPADES/Output/misc/assembled_contigs.fasta']" finished abnormally, err code: 255

params.txt spades.log

[asl]

Hello

Thank you for your interest in SPAdes.

I got "Segmentation fault (core dumped)" message at bwa mem alignment step (the last few lines copied below my message and attached are params.txt and spades.log). Is this a memory issue or any issue due to read files being .fasta files? The genome I'm trying to assemble is 1.6 Gbp.

No, it's bwa being unable to handle input with reference size more than 2Gb. We will likely fix this in the next release.

Do the steps starting from bwa mem (polishing etc) improve the assembly statistics (N50 etc.) any further or do the files assembled_contigs.fasta and assembled_scaffolds.fasta represent the last stage of 'assembly'?

You may expect smaller number of mismatches / short indels. Though no n50 improvements will be there :)

[githubgig]

Thanks for the quick response. I guess I'll make do with assembled_scaffolds.fasta for now.

[halwap]

Hi,

I have very similar (if not the same issue).

0:37:14.967 1G / 1G INFO DatasetProcessor (dataset_processor.cpp : 149) Running bwa mem ...:/home/halakuc/software/SPAdes-3.13.1-Linux/bin/spades-bwa mem -v 1 -t 24 /home/halakuc/genomy/egracilis/spades_miseq_hiseq/misc/assembled_contigs.fasta /home/halakuc/genomy/egracilis/Egracilis_HiSeq_norm_prefilter_1.fq /home/halakuc/genomy/egracilis/Egracilis_HiSeq_norm_prefilter_2.fq > /home/halakuc/genomy/egracilis/spades_miseq_hiseq/tmp/corrector_6Wf8oI/lib0_ZYWQOO/tmp.sam Segmentation fault (core dumped) 0:37:22.836 1G / 1G INFO DatasetProcessor (dataset_processor.cpp : 152) bwa failed, skipping sublib 0:37:22.836 1G / 1G ERROR DatasetProcessor (dataset_processor.cpp : 215) Failed to align paired reads /home/halakuc/genomy/egracilis/Egracilis_HiSeq_norm_prefilter_1.fq and /home/halakuc/genomy/egracilis/Egracilis_HiSeq_norm_prefilter_2.fq

I'm assembling HiSeq + MiSeq data, and assembled_contigs.fasta is 1.4GB. Project I'm working on focuses on atypical introns, so I would much prefer working with corrected scaffolds. Is there anything I can try to do to correct them? Is there possiblity that pre-release SPAdes 3.14 will work or was it not touched yet?

[asl]

So far the only workaround is to recompile SPAdes from source. Will be an option for you?

[halwap]

Probably yes. Which version should I use? Any special steps too follow?

[asl]

Just take the latest released version. You need to modify https://github.com/ablab/spades/blob/spades_3.13.1/assembler/src/projects/corrector/dataset_processor.cpp#L139

Just change "is" to "auto" there and recompile.