Increase memory requirements for STAR genome generate

[pfgherardini]

Description of feature

I think the STAR genome generate process should be labeled as process_high_memory. The current defaults are not enough to index the human genome even on the second attempt

[drpatelh]

Hi @pfgherardini ! It would be a good to get a gauge of exactly how many resources are required to index the human genome? We can attempt to bump the resources based on that.

[drpatelh]

Will close this for now as I suspect the pipeline is being used routinely on human genomes. Be good to know what sort of memory requirements we need to bump to. Please feel free to re-open if the problem persists or if we can use any new info regarding memory requirements.