Bowtie2 for miRNA-Seq

[gkuffel]

Hopefully someone can help. I have read conflicting info about aligners for miRNA data but I have finally decided on Bowtie2. I have 3 questions:

1. Can anyone share the parameters they are using to run Bowtie2 to align miRNA and give a short explanation for the choices?

2. I am aligning to the human genome and then using HTSeq with the gff file from miRBase to count the miRNAs aligned in the BAM file, is this a reasonable approach?

3. Does anyone know if HTSeq will accurately count miRNAs that map to multiple locations in the genome?

Thanks everyone!

[sagarutturkar]

I think bowtie is an appropriate choice for mapping the miRNA reads. Bowtie is more sensitive for the reads < 50 bp. Standard miRNA analysis software mirdeep2/mirExpress are based on bowtie aligner.

I would definitely keep the number of mismatches allowed = 0.

You may want to compare the mapping results with bowtie/bowtie2. Bowtie2 might generate higher mapping rate. But bowtie2 is more lenient and can also include ambiguously mapping reads.

[dhaithry thrayambak]

You can get clear idea about these in Bowtie's tutorial page.

It is good to align miRNA sequence.
your second one is surely a reasonable approach.
Accuracy varies but you can expect the aligned ones are good