Parameters to consider for aligning long reads (150bp) in Tophat

[Studentlost]

Tophat's defaults sets: --read-mismatches 2 --read-gap-length 2 --read-edit-dist 2.

This gave me about 4% alignment for mm9 mouse data.

I tried up to --read-mismatches 30 --read-gap-length 30 --read-edit-dist 30. This gave me 20.9% alignment.

The reads are 150 bp long. What are the proper settings for such long reads? If I wanted to use tophat, rather than other tools that may be better suited for long reads, what would I need to take into consideration? Any other parameters?

[Brian Bushnell]

You may have low-quality data, or short-insert data, in which cases you may want to do quality or adapter trimming first to improve the mapping rate. If that does not help - I am not an expert on Tophat's parameter space, but I am for BBMap, which typically yields a higher mapping rate.