Hi, I am trying to map some short RNA-Seq reads from mouse against genome (mm10)/transcriptome sequences using bowtie2. There is a particular case which I couldn't resolve. Briefly, bowtie2 is not able to find an alignment for a read of length 33 which in fact can align against the genomic sequence with a single gap roughly in the middle of the read sequence.
When this gap is corrected by hand, then bowtie2 finds the match. To be more specific the particular sequence is from a U2 ncRNA and a tiny fastq format is attached. The second read in the same file is a faux read created by changing the ID of the first one and filling in the gap by hand.
I have tried bowtie2 with several parameters:
--very-sensitive
-L 10 -D 20 -R 20 -N 0 -i S,1,0.5
-L 10 -D 20 -R 20 -N 1 -i S,1,0.5
-L 10 -D 20 -R 20 -N 0 -i S,1,0.5 --rdg 1,1
-L 6 --score-min=C,-60,0
-L 6 --score-min=C,-100,0
Only in the last one, I started to get some horribly bad misalignments, but not the aimed alignment with one gap. Can someone explain what I am missing? Thanks for any comments/suggestions.
HTML Code:
Query 1 ATACGTCCTCTATC-GAGGACAATATATTAAATG 33 |||||||||||||| ||||||||||||||||||| Sbjct 64 ATACGTCCTCTATCCGAGGACAATATATTAAATG 97
I have tried bowtie2 with several parameters:
--very-sensitive
-L 10 -D 20 -R 20 -N 0 -i S,1,0.5
-L 10 -D 20 -R 20 -N 1 -i S,1,0.5
-L 10 -D 20 -R 20 -N 0 -i S,1,0.5 --rdg 1,1
-L 6 --score-min=C,-60,0
-L 6 --score-min=C,-100,0
Only in the last one, I started to get some horribly bad misalignments, but not the aimed alignment with one gap. Can someone explain what I am missing? Thanks for any comments/suggestions.