hisat2: "(ERR): hisat2-align died with signal 6 (ABRT) "-message

Question

[Ad fontes]

Hi,

I am completely new to this, any help will be greatly appreciated.

After running hisat2, I get

“Error, fewer reads in file specified with -1 than in file specified with -2
libc++abi.dylib: terminating with uncaught exception of type int
(ERR): hisat2-align died with signal 6 (ABRT) “

What does this mean?

I am trying to match two sets of single-end reads against the reference genome (taken from ensemble)

I have first built an index:

hisat2-build someSpecies.toplevel.fa   someSpeciesIndex

After extracting splice-sites:

hisat2_extract_splice_sites.py someSpecies.gtf > someSpecies.txt

I run hisat2:

hisat2 -x  someSpeciesIndex  --known-splicesite-infile     someSpecies.txt  -p 3 -1 ourRead1.fq   -2 ourRead2.fq  > mapped_reads.sam

Then I get the above message. The mapped_reads.sam =10,36GB

What am I doing wrong?

[Ad fontes]

It is single-read, but even when I specify rna-strandness R I get the same answer. Any help would be greatly appreciated

[GenoMax]

Why are you using the option for paired-end reads if you have single-end data?

[Ad fontes]

According to https://daehwankimlab.github.io/hisat2/manual/ we should use --rna-strandness F (or R) if we have single-end reads?

[GenoMax]

RNA-strandedness does not per se have anything to do with single-end reads. You need to know if the libraries you are working with were prepared using a standed protocol (which preserves information about the strand of the DNA the RNA came from).

[GenoMax]

Looks like you don't have the same number of reads in Read 1 and Read 2 file. This can happen if you did your scanning/trimming independently. You should always trim paired end reads together to avoid getting reads out of sync.

You can use repair.sh from BBMap suite to bring your paired-end reads back in sync and remove any singletons. A guide is available here.

[Ad fontes]

Sorry, I don't think this is the answer: I am trying to match two sets of reads against the reference genome: why should it matter if I don't have the same number of reads in Read 1 and Read 2?

Answer 1

[GenoMax]

If you provide two different sets of reads with -1 and -2 option the software thinks that they are from a paired-end dataset. In that case the number of reads needs to match since every fragment in your library produced the two reads from two ends.

If these are two separate samples then you should run them using 2 independent alignment jobs using option for single end reads.

[Ad fontes]

Ah, thanks, that makes sense!