Metadata
| Status | done |
|---|---|
| Assigned | agent-2707 |
| Agent identity | 3184716484e6f0ea08bb13539daf07686ee79d440505f1fdf2de0357707034c3 |
| Created | 2026-06-23T16:01:30.834619188+00:00 |
| Started | 2026-06-24T07:29:49.205622075+00:00 |
| Completed | 2026-06-24T07:35:37.559652407+00:00 |
| Tags | fig5, figure, query-grid, sweepga, eval-scheduled |
Description
Regenerate the Fig5 raw FASTA SweepGA f16 panels using the query-grid chopped/filter outputs.
Goal: replace/augment the current row-start-chopped sensitivity view with a query-grid version where every mapping is cut on absolute query-coordinate boundaries before 1:1 filtering.
Required outputs:
- A new figure package under paper_prep/_brainstorming/fig5_raw_fasta_sweepga_f16_query_grid_chop_filter_panels/ or an equivalently explicit query-grid directory name.
- PDF, PNG, and SVG panel outputs.
- Segment-level TSV and summary TSV with genomic query coordinates, not only 500 kb window coordinates.
- Manifest that records raw many:many source PAFs, query-grid chopped PAFs, filtered PAFs, SweepGA command settings, pafchop command settings, and sha256 checksums.
Plot requirements:
- Show the PAR1 control plus the PAN027 and PAN028 chr9q -> chr3q candidate windows.
- Include the query-grid chop lengths produced by the rerun, at minimum 10 kb, 5 kb, 2 kb; include 1 kb if available.
- Clearly label query-grid chopping and SweepGA --overlap 0 / --scoring ani / --scaffold-jump 0.
- Keep raw many:many support visually separate from post-chop/filter 1:1 support.
- Fix the confusing legend/overprint seen in prior panels.
Acceptance criteria:
- Figure outputs render and validation script passes.
- Summaries report chr3 retained union bp and target-chrom union bp by candidate, chop length, and filter mode.
- The README explains why query-grid chopping differs from the older row-start chopped panels.
- Commit message follows repo convention: feat: fig5-query-grid-panels-regenerate (agent-NNN)
Depends on
Required by
Log
- 2026-06-24T07:29:49.205625622+00:00 Spawned by coordinator --executor codex --model gpt-5.5
- 2026-06-24T07:29:58.754356070+00:00 Evaluator starting: inspecting task status, artifacts, commit, and validation evidence
- 2026-06-24T07:32:56.679138655+00:00 Evaluation drafted: no query-grid panel package or task-specific commit found; provisional grade 0.00
- 2026-06-24T07:33:47.718722297+00:00 Validated evaluation: searched for query-grid package, rendered outputs, TSVs, README, manifest, and task-specific commit; none found; grade 0.00 recorded
- 2026-06-24T07:34:18.607047137+00:00 Committed evaluation: bd132aa
- 2026-06-24T07:35:02.668735182+00:00 Committed and pushed: bd132aa — eval: fig5-query-grid-panels-regenerate (agent-2707)
- 2026-06-24T07:35:37.559659982+00:00 Task pending eval (agent reported done; awaiting `.evaluate-*` to score)
- 2026-06-24T07:47:34.284675537+00:00 PendingEval → Done (evaluator passed; downstream unblocks)