long read files input format problem

[87joe]

I used CONSENT v2.0 to do nanopore long reads correction, but I encountered some problem all errors are core dump.
I traced the source code and found the function indexReads in utils.cpp is the problem.
This function read readsfile in the pattern "1 header 1 seq", e.g,
>seq001
TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTAGGAAATAAAGTAAATTTTTGTTGTTGTACTTCGTTCAGTTTGGGTGTTTAACCAGATGTCGCCTACCGTGACAAGAAAGTTGAAAGAAAATAAGAAAATACGGCGCTGTCGCGGTTCGAACCACAGACCTTGACCCCCAGCAATATCAGCACCAACGAAACACAAGACACCGACAACTTTCTTGTC
So I modified my fasta format like the pattern, CONSENT worked.
My original fasta file format(60 char per line every sequence):
>seq001
TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTAGGAAATAAAGTAAATTTTTGTT
GTTGTACTTCGTTCAGTTTGGGTGTTTAACCAGATGTCGCCTACCGTGACAAGAAAGTTG
AAAGAAAATAAGAAAATACGGCGCTGTCGCGGTTCGAACCACAGACCTTGACCCCCAGCA
ATATCAGCACCAACGAAACACAAGACACCGACAACTTTCTTGTC
But when I correcting another nanopore long reads file it was a new problem minimap2 core dump.
I tried to run minimap2 with original format was working, but the CONSENT will core dump
first file is 1G, modified format all good
second file is 13G, modified format minimap2 core dump

[morispi]

Hi,

This a weird error, I never encountered any issue with Minimap2 and 1 header 1 seq reads file. Did you try to launch only Minimap2 on your 13G long reads file and see what happened?

Also, does your file contain extremely long or extremely short reads? That'd be weird, but maybe it could cause the issue.