Sbp-359 Optimise gadi config - wisps

[Mitchob]

updates

• Increase batch sizes based on Wenjing benchmarking suggestions
• Cleanup intermediate files with cleanup. Spoke to Tom about using nf-boost agreed to leave it at this time to avoid adding another plugin to manage
• added scratch = true for RUN_BOLTZ

cleanup

Cleaned up wisps.config by removing settings that duplicated defaults, the file should now only contain genuine NCI/SBP-specific overrides.

more explicitly I cleaned up:

• Redundant params that match defaults (use_gpu, colabfold_num_recycles, ipsae_pae_cutoff/dist_cutoff, colabfold_uniref30_prefix/envdb_prefix)
• 4 unused params (boltz_ccd, boltz_model, uniref30, colabfold_db) not referenced in the pipeline and were giving warnings
• Generic process resources and all withLabel: blocks (identical to base.config)
• shell, ext.args, singularity.registry/runOptions, nextflow.enable.configProcessNamesValidation, and the manifest {} block, and validation.defaultIgnoreParams

This should have no functional changes.

[eileen-xue]

Some GPU tasks requesting 1–2 days of runtime. If the task doesn't need that long, request less time — dgxa100 and gpuhopper nodes are always busy, so shorter requests mean shorter queue waits.

[Mitchob]

@eileen-xue ready for a second review. should mem for hugemembw be 584 or 585 GB?

[eileen-xue]

512GB/14*16=585.14GB.