Solved: BWA ALN cannot not find a very obvious match

Question

[max]

(Originally from Milan Simonovic posted on sourceforge but I could reproduce it and have the same problem, https://sourceforge.net/p/bio-bwa/mailman/message/36190558/)

We have BWA at full sensitivity here and it still cannot find a very obvious sequence in zebrafish chr10. Any ideas what we're missing?

# cat reads.fa
> testSeq
TTTATTTCCACACTTCATGG

Summary of alignable sequences on chr10:

TTTATTTCCACACTTCATGG chr10:10367145-10367165 + this is the test sequence
TTTATTTCCACATATGATGG chr10:25378199-25378221 - this is the sequence that was not found
TTTATTTCCACA**T*ATGG

steps to reproduce:

wget -O chr10.fa.gz ftp://ftp.ensembl.org/pub/release-85/fasta/danio_rerio/dna/Danio_rerio.GRCz10.dna.chromosome.10.fa.gz

$ ./bwa

Program: bwa (alignment via Burrows-Wheeler transformation)
Version: 0.7.17-r1188
..
$ ./bwa index chr10.fa.gz
$ ./bwa aln -o 0 -n 4 -k 4 -N -l 20 chr10.fa.gz reads.txt > reads.sai
$ ./bwa samse -n 4 chr10.fa.gz reads.sai reads.txt
# returns just the exact match
[bwa_aln_core] print alignments...  0   10  10367146    2   20M *   0   0   TTTATTTCCACACTTCATGG    *   XT:A:U  NM:i:0  X0:i:1  X1:i:135    XM:i:0  XO:i:0  XG:i:0  MD:Z:20

Answer 1

[max]

Sorry, this was stupid. I have to run bwa samse like this to get all matches: "bwa samse -n 100000000". Never mind.