helixforge parallel aggregate
Collect per-chunk outputs by pattern → one annotation; verify + fail closed.
Step 4 of the chunked path. Merges the per-chunk outputs (by --input-dir +
--pattern, or explicit --chunk-output prefixes) into one genome-wide
annotation and folds the chunk id_maps into --master-id-map. The merge is
verifying: it asserts globally-unique gene AND transcript ids and that every
Helixer locus is covered exactly once, raising on any violation.
Outputs (under the resolved prefix):
.gff3 (+ .tier1/2/3.gff3) and .report.tsv, the merged annotation + per-gene report (same columns as reconcile)- the master id_map.json (created if absent)
- a printed summary: num_genes, num_loci, tier_counts, origin_counts
Options
Section titled “Options”| Option | Description |
|---|---|
--input-dir | Directory of per-chunk outputs (use with --pattern). |
--pattern | Glob for the per-chunk GFF3s under --input-dir. (default: *.gff3) |
--chunk-output | Explicit per-chunk output prefix (repeatable; alternative to --input-dir). (repeatable) |
-o, --output | Merged GFF3 path, e.g. combined.gff3 ('.gff3' stripped for the prefix). |
--out-prefix | Output prefix (alternative to -o/--output). |
--master-id-map | Master id_map.json to fold chunk maps into (created if absent). |
Examples
Section titled “Examples” helixforge parallel aggregate --input-dir chunks/ --pattern '*.gff3' \ -o maize_helixforge.gff3 --master-id-map master_id_map.json