Abyss-PE error

[joa_ds]

Hi community,

I am currently involved in a rather complex project where we are hunting for causal SNPs in a species that has no assembled genome at the moment. We are planning on sequencing some normal plants to get partial assemblies and then in another run a pool of inbred mutants and hopefully pick up the causal mutant.

Anywayz, needless to say, the assembly is crucial. So I am using Abyss in an experimental phase to test different settings (>1000 settings to test...). But before I start 1000 assembly jobs, I need to be sure that Abyss is working as desired.

I just take a genome of a related species, chop it up, get some errors in, pick a coverage, feed it to abyss and check for avg contig size, size distribution, % of original covered etc.

regular ABYSS is working like a charm.
abyss-pe is not...

My fasta file looks like reads_1.fa
>1A
ACGACTACGCGA
>2A
ACGACTACG

reads_2.fa
>1B
ACGACGA
>2B
ACGACTCAG

The commando I run:
/data/abyss-1.2.5/bin/abyss-pe k=15 n=40 -o 3334.assembly.fa in='3543.reads_1.fa 3543.reads_2.fa' name=3334

The error I get:
ABYSS -k15 -q3 --coverage-hist=coverage.hist -s 3334-bubbles.fa -o 3334-1.fa 3543.reads_1.fa 3543.reads_2.fa
ABySS 1.2.5
ABYSS -k15 -q3 --coverage-hist=coverage.hist -s 3334-bubbles.fa -o 3334-1.fa 3543.reads_1.fa 3543.reads_2.fa
Reading `3543.reads_1.fa'
ABYSS: FastaReader.cpp:85: bool isColourSpace(const std::string&): Assertion `!seq.empty()' failed.



What am I doing wrong?

[peromhc]

get rid of that

-o 3334.assembly.fa

part of your command and you should be OK.