Transdecoder - log likelihood for which purpose ?

Question

[Juke34]

I don't get the use of log-likelihood score by Transdecoder. I googled and didn't find satisfying information. Someone could provide some hints ?

Why using likelihood when searching longest ORF ? I have the feeling it assumes that some frames are more likely than some others... but why ? In a transcritpome we don't know the frame...

Answer 1

[Juke34]

I found my answer in the supplement of the trinity paper here