Dynamic scaling of compute specifications for GeneRax

[austinhpatton]

Description of feature

Although the process specification for GeneRax works for the vast majority of gene families, there will always be a small number of significantly larger gene families for which the default process/compute specifications are insufficient. This can lead to significant delays in workflow completion, as these large gene families will fail repeatedly (after all smaller gene families have finished) until appropriate resources are allocated.

Ideally, we could bin gene families according to size (# species, # proteins, length/complexity of sequences), and allocate to each bin sensibly scaled compute specifications.