Bring usegalaxy.org rule for hisat2 to the shared TPV database

[kysrpex]

Hi! I was looking into an OOM issue with HISAT2 on usegalaxy.eu and although it's unrelated, at some point @bgruening mentioned the TPV rule they have at usegalaxy.org for HISAT2.

It looks better than what the have in the shared DB, it seems to provide more breathing room when processing large datasets with HISAT2. However I don't know how well it works, I guess @natefoo can provide better insights and help us decide if the switch is worth it.

[nuwang]

I see the ranges on Galaxy Australia are somewhat different: https://github.com/usegalaxy-au/infrastructure/blob/fdc808f754deb1f328b4d4e73b0148be3ed36b13/files/galaxy/dynamic_job_rules/production/total_perspective_vortex/tools.yml#L2009
Do you want to comment here @cat-bro?

[cat-bro]

LGTM. It’s not that different from AU’s set of rules except for the largest allocation. We were seeing OOM errors for hisat2 and switched to giving the big jobs cores: 36, mem: 500 over a year ago. Some jobs run out of memory even with this allocation. It would be interesting to understand the parameters that lead to such high memory use, because 500GB is excessive for most of the jobs.

[nuwang]

Thank you @kysrpex!