BWA index segmentation fault for specific sequence

Question

[F Lekschas]

I am experiencing a strange error with bwa index. I want to index the KoRV fasta file but bwa index keeps running into a segmentation fault. What is most confusing is that when I delete line 117 from the fasta file (given the same line length as from genbank) then bwa index works like a charm!?!

I am using Mac OS X 10.8.5 with bwa 0.7.7 and 0.7.8-r455. And bwa worked with any other (and much bigger) fasta files.

Has anyone an idea of this is a bug of bwa or if I am making somethig wrong?

UPDATE:

I figured out with works without -a bwtsw and using the default is algorithm.

Answer 1

[Dan D]

There's nothing about the text fasta version of the genome that would prevent BWA from using it as-is.

With bwa 0.7.7:

$ bwa index -a is korv.fa
[bwa_index] Pack FASTA... 0.00 sec
[bwa_index] Construct BWT for the packed sequence...
[bwa_index] 0.00 seconds elapse.
[bwa_index] Update BWT... 0.00 sec
[bwa_index] Pack forward-only FASTA... 0.00 sec
[bwa_index] Construct SA from BWT and Occ... 0.00 sec
[main] Version: 0.7.7
[main] CMD: bwa index -a is korv.fa
[main] Real time: 0.014 sec; CPU: 0.006 sec

You should not use bwtsw as your algorithm, as it's not for short genomes (and this is a very short genome). If I understand your comment correctly it sounds like you might have figured that out, but it's not totally clear.

[F Lekschas]

Yes I figured out it works with -a is. But thanks for the explanation that bwtsw does not work with small genomes. I didn't know about that.