hisat2-align died with signal 11 about hisat2

[homopolymer]

I ran make_grcm38_tran.sh to re-build the index. But it did not help.

[filicado]

Same here. It works fine with the majority of paired end fastq files using the same index but dies with some with that error.
For a start the error message isn't very informative, it would help to know was being aligned when this happens.

[filicado]

Should add: the same fastq files align just fine with the original hisat

[infphilo]

Sorry for this problem - would you guys be able to let me know your specific HISAT2 commands? I'll do testing on my end and fix the problem.

[filicado]

Pretty standard command aligning paired end fastq files:

perl /media/BioTools/hisat2_v2.0.0/hisat2 -x /media/BioProjects/Reference/HISATRef -p 8 -1 /media/BioProjects/Reads/Female_R1.fastq -2 /media/BioProjects/Reads/Female_R2.fastq | /media/BioTools/samtools_v0.1.19/samtools view -bhSu - | /media/BioTools/samtools_v0.1.19/samtools sort -@ 8 - /media/BioProjects/BAMs/Female.sorted

(I slightly shortened the paths to make this readable)
But as I said the command works fine with the original hisat and it works fine with most samples even with hisat2. So there must be something about specific fastq files that it doesn't like.

[infphilo]

This problem might have been fixed in the GitHub version of HISAT2. Otherwise, I'd like to run HISAT2 on one of your data sets to investigate what goes wrong, if possible. Please email me ([email protected]) if you need a FTP site where you want to upload your data.

[infphilo]

@filicado Thank you for providing your data set, I just fixed this problem, which happened when reads map beyond the first base of a reference sequence.. The fixed version will be released within two weeks. , or you can use the GitHub version of HISAT2.

[infphilo]

@homopolymer I also fixed this issue (it's a bug different from what@filicado reported). It's due to illegal memory access. The fixed version will be available within two weeks.