Support interleaved paired-end reads [feature request]

[sjackman]

Hi, Ryan. Please consider supporting assembling interleaved paired-end reads in one FASTQ file. Thanks!

[sjackman]

As a workaround…

seqtk seq -1 foo.fq.gz | pigz -p64 >foo.1.fq.gz
seqtk seq -2 foo.fq.gz | pigz -p64 >foo.2.fq.gz
unicycler -t64 -o foo.unicycler -1 foo.1.fq.gz -2 foo.2.fq.gz

# Separate the first read from an interleaved FASTQ file.
%.1.fq.gz: %.fq.gz
	seqtk seq -1 $< | pigz -p$t >$@

# Separate the second read from an interleaved FASTQ file.
%.2.fq.gz: %.fq.gz
	seqtk seq -2 $< | pigz -p$t >$@

# Assemble reads using Unicycler.
%.unicycler/assembly.fasta: %.1.fq.gz %.2.fq.gz
	unicycler -t$t -o $(@D) -1 $*.1.fq.gz -2 $*.2.fq.gz

[rrwick]

Thanks - I'll add this to my future feature list. In the meantime, I appreciate the workaround!

[sjackman]

If you do, please consider implementing the same smart pairing feature as bwa mem -p, which allows for unpaired reads in the same interleaved paired-end file.

❯❯❯ bwa mem |& grep -- -p
       -p            smart pairing (ignoring in2.fq)

[sjackman]

As a command line interface, I'd suggest…

unicycler -o output [file…]

where [file…] is one or more interleaved FASTA/FASTQ possibly compressed files.

[sjackman]

I'd still love to see support for interleaved FASTQ files for unicycler and unicycler_polish.

[rhysnewell]

Hi Ryan,

Just wondering if there was any plan to implement this sometime soon or if it has been implemented but is just undocumented?

Cheers