helixforge parallel plan

Partition the genome (v1 strategies) + reserve disjoint HFG ranges → plan.json.

Step 1 of the chunked path (plan → tasks → run → aggregate). Cuts the genome only in inter-locus gaps wider than --min-boundary-gap so no gene, and no Mikado merge, is ever split, then reserves each chunk a disjoint, contiguous HFG number range (seeded from --id-map for run-stable ids). The plan is consumed by parallel tasks.

Output: plan.json with top-level {strategy, min_boundary_gap, flank, total_loci} and a "chunks" list; each chunk has: chunk_id, regions, num_loci, locus_ids, id_base, id_range, novel_base, novel_range, est_resources.

Options

Option	Description
--genome	Genome FASTA or .fai index (for scaffold sizes). (required)
--helixer, --gff	Helixer GFF3 (master loci; drives gene-aware cuts + id ranges). (required)
--strategy	Chunking strategy (v1 vocabulary). (choices: scaffold/size/genes/adaptive; default: scaffold)
--chunk-size	Bases per chunk for 'size'; loci per chunk for 'genes'.
--min-chunk-size	Smallest chunk (bp) for the size-based strategies. (default: 100000)
--max-chunk-size	Split a scaffold longer than this (bp) under 'scaffold'.
--target-chunks	Target chunk count for 'adaptive' (from parallel suggest).
--min-boundary-gap	Only cut in inter-locus gaps at least this wide (bp), so a gene is never split. Default max(--flank, 1000).
--flank	(default: 200)
--id-map	Master id_map.json to seed stable HFG ranges from.
-o, --output	Output plan path. (default: plan.json)

Examples

  # one chunk per scaffold (default), stable ids seeded from a prior run
  helixforge parallel plan --genome maize.fa --gff helixer.gff3 \
      --id-map master_id_map.json -o plan.json

  # ~2000 loci per chunk
  helixforge parallel plan --genome maize.fa --gff helixer.gff3 \
      --strategy genes --chunk-size 2000 -o plan.json