Description of feature
It would be great if the pipeline split each of the Trinity stages (jellyfish, inchworm, chrysalis, butterfly) into distinct processes so they could be better targeted to cpu and memory resources required for each task and make more efficient use of HPC (and Batch etc) resources.
An example implementation of this for reference would be: https://github.com/marcodelapierre/trinity-nf