alignment to human transcriptome with Bowtie2

Question

[Meghan.T]

I have a general question. So I see that in some publications when analyzing RNA-seq data they use Bowtie2 and human transcriptome. As I understand, Bowtie2 can not detect splice juctions, but I assume that this information is readily available in the human transcriptome. I usually use HISAT2 for RNA-seq alignment and I was wondering besides identifying alternative splicing events, Is there any advantage or difference between HISAT2 and Bowtie2+human transcriptome?

Answer 1

[swbarnes2]

I think it's generally preferable to use a splice-aware aligner to align to the whole genome, not to use a non-splice aware aligner to align to a multi-fasta of transcripts. So I would say keep doing what you are doing.

[dsull]

Hmm, I wouldn’t say that’s generally true. Good answer here: https://www.reddit.com/r/bioinformatics/comments/191bc7n/spliceaware_vs_nonaware_aligners_genelevel_vs/

[swbarnes2]

There are of course tools designed to give accurate transcript level counts by aligning to a file of transcripts, bowtie2 is not one of them.