Closed lucventurini closed 8 years ago
Should be fixed in the latest version of the development branch. However, I couldn't fully test this out this morning as SLURM wasn't allowing jobs to run correctly. We will need to test properly later today.
Fixed in 1.0.0b3.
At the moment Daijin executes TransDecoder without asking for the minimum length of the ORF to be found. The default, 100 AA or 300 bps, is too high for Mikado as the tool relies on the coding signal for many purposes. I suspect that many of the issues I have with the simulated data derive from the setting being too high; there are thousands of reconstructable transcripts with an ORF shorter than 100 AA, which we lose as putative fragments.