Bowtie2 mapping of (several, long) contigs in fasta to a reference sequence - contig sequences only mapped as 60 bp

Question

[kalfsnes]

I've got contigs from de novo (Spades) assembly (approx 175 contigs of length 200 bp to 11 kb), and want to map these to a reference sequence and make a new consensus. I have tried using Bowtie2 and I do get an output. However, all the aligned contigs have been truncated down to 60 bp.

I've tried changing the Bowtie2 index to "large index", Bowtie2 is run using the -f and -U options.

Any help appreciated.

Answer 1

[Devon Ryan]

1. Does your fasta file have 60 base wide lines?
2. You don't need a large index.
3. I would strongly recommend that you use bwa (or minimap2) instead. It's unclear how good bowtie2 will be with long contigs.

[kalfsnes]

1. Yes, but shouldn't Bowtie2 handle these line breaks? Anyway, I removed them, but that left me with a new problem - memory. I'll try run it on computer with more memory, see if that solves it.
2. Okay.
3. Worked like a charm with bwa - thanks! :-)

[Devon Ryan]

Nah, it expects sequences on a single line for fastq files, so fasta will be the same. I expect most aligners that accept fasta input are like that.