Adherence to the "-G" parameter

[sanyalab]

Hello,

In minimap the -G parameter decides the maximum gap on the reference during spliced alignments. So I take this as either intron length or intergenic length. Is my assumption correct? This is set to 200000 by default. I set the -G value to 50000 for my species, as the 90th percentile value of intron length was 48000. It turns out that the gene I was interested in had a single 60000 bp intron. Therefore, FLNC reads that should have aligned did not, as I set -G 50000. When I ran minimap with default -G the reads aligned fine. Therefore, a few questions.

1. Is it advisable to reduce this size closer to the anticipated intron size of a species?
2. How strictly does the tool adhere to the -G parameter value? You mention that this has an effect on the chaining and alignment band width. Which are these parameters? and what do they control?

Thanks
Abhijit

[sanyalab]

Hi @lh3

I was wondering if you had answered the question above as part of a different thread. Could you please provide the link to that thread? I'll look through it. I could not find an answer therefore asking.

Thanks
Abhijit