helixforge parallel suggest

Recommend chunk count + per-chunk resources (heuristic; prints trade-offs).

Step 0 of the chunked path. From the genome size/scaffold profile and your node spec (--cores-per-node, --mem-per-node, --max-array-size, --walltime-cap) it suggests a granularity and per-chunk resources to feed into parallel plan / run --scatter. Heuristic only, it reads sizes, runs nothing, and modifies nothing.

Output (printed recommendation): target_chunks, target_loci_per_chunk, min_boundary_gap, and per-chunk procs / threads / mem_gb / walltime_min, with a rationale explaining the trade-offs.

Options

Option	Description
--genome	Genome FASTA or .fai index. (required)
--cores-per-node	(default: 16)
--mem-per-node	RAM per node, GB. (default: 64)
--max-array-size	(default: 1000)
--walltime-cap	Walltime cap, hours. (default: 24)

Examples

  # default node profile
  helixforge parallel suggest --genome maize.fa

  # tune to a specific node + array/walltime caps
  helixforge parallel suggest --genome maize.fa --cores-per-node 32 \
      --mem-per-node 128 --max-array-size 500 --walltime-cap 12