search

pmelsted / pizzly

Fast fusion detection using kallisto
BSD 2-Clause "Simplified" License
80 stars 10 forks source link

Max insert size from Kallisto #45

Closed MikeWLloyd closed 1 year ago

MikeWLloyd commented 1 year ago

I found issue #5, and the final comment quoted below. The question posed was never answered, and I am also curious about the answer.

Should I use mean, 95%, or 'max' insert size? If max, what would the calculation be? In #5 there is an argmax statement, would I simply remove '> .95', or would the max calculation be something else? I can't follow what is stored in the matrices, and would appreciate any help. 

          @pmelsted in the command line args it says the "max" value, but do you suggest to instead use the 95th percentile as shown in this script? It is very different from the max value so I wanted to get your input.

Originally posted by @kmhernan in https://github.com/pmelsted/pizzly/issues/5#issuecomment-315822547

pmelsted commented 1 year ago

The max parameter in pizzly only uses read pairs where the insert size is below the cutoff. My suggestion is to use the 95% of the observed distribution of insert sizes where both reads map concordantly as the cutoff.

MikeWLloyd commented 1 year ago

Thank you very much for the clarification! I will move forward with 95%.