Bowtie2 mismatches option for small RNA mapping

[jazz]

I am using Bowtie2 to map small RNA reads (not microRNAs) to the genome. To account for allelic divergence, I need to set it to allow 1 or maybe even 2 mismatches. Is it possible to change that? I believe the N -1 option only lets you allow mismatches in the seed region.
I am using end-to-end very sensitive option.

Thanks for any insights.

[Kennels]

I think bowtie1 would be better suited for your purposes. I use bowtie1 for small RNA alignments. Or Patman is another.

[jazz]

But bowtie1 does not allow for gapped alignment and I am expecting some possible small RNAs coming from fusion/chimeras. I am thinking about Blat and Tophat fusion. Any suggestions on those?

[Kennels]

Ah sorry I misread the part that you are not looking for miRNAs.
Patman handles gaps too, though it becomes slower.

[joseph.troy]

I spent yesterday reading the bowtie 2 manual trying to figure out how to limit the output to alignments with no more than 1 mismatch. (in bowtie 1 I was able to limit output to those with no more than one mismatch using the -v 1 option at the command line).

It seems in bowtie2 you must filter out alignments with more than 1 mismatch (or y mismatches) AFTER the sam output file is produced. The bowtie 2 manual describes "optional" fields in the sam output. One of those optional fields is the XM:i:<N> flag which is "The number of mismatches in the alignment. Only present if SAM record is for an aligned read." Below is partial SAM output showing those flags. (the below just shows part of the rows with some of the flags) The output shows an alignment with 7 mismatches, then one with 1 mismatch and then one with 0 mismatches.


GGGGFGGEEGEEG AS:i:-34 XS:i:-44 XN:i:0 XM:i:7 XO:i:0 XG:i:0 NM:i:7 MD:Z:9A2G0T3A2A7C6A39 YT:Z:UU
EDEBD@DEEEEB@@ AS:i:-5 XN:i:0 XM:i:1 XO:i:0 XG:i:0 NM:i:1 MD:Z:21A53 YT:Z:UU
FEEFBFFBB AS:i:0 XN:i:0 XM:i:0 XO:i:0 XG:i:0 NM:i:0 MD:Z:75 YT:Z:UU

So, my plan is to write a perl script to remove those with more than one mismatch before I continue my downstream processing of the data.

Any thoughts or comments would be appreciated, I'm just trying to learn how all this works as well.

-Joe