Helix StudioGenome IDE
← Docs
Helix CLI docs
Browse Helix CLI docs

Edit DAG Visualization Toolkit

Helix emits deterministic DAG artifacts (helix.crispr.edit_dag.v*, helix.prime.edit_dag.v*, helix.pcr.amplicon_dag.v*). This page documents the premium rendering tools that turn those JSON artifacts into publication-ready figures, GIFs, and shareable reports.

Rendering options

CLI PNG render

helix edit-dag viz --input dag.json --out dag.png \
  --layout timeline \
  --min-prob 0.02 \
  --max-time 8
  • --layout: timeline (default) for left→right stages, or spring, kamada-kawai, etc. for exploratory plots.
  • --min-prob, --max-time: filter low-probability branches or truncate long simulations before plotting.

Output:

  • dag.png — cinematic lineage figure (stage-colored rings, probability-weighted edges, top-outcome inset).
  • dag.viz.json — viz-spec with schema metadata.
  • dag.provenance.json — cryptographic audit trail (schema kind, spec version, checksums, CLI command).

CLI GIF animation

helix edit-dag animate --input dag.json --out dag.gif \
  --layout timeline --fps 6 \
  --min-prob 0.02 --max-time 8

Produces a frame-per-time-step GIF with glowing highlights for emerging branches, temporal bands, and probability-aware arrow widths.

Playground

https://omniscoder.github.io/Helix/playground/

  • Drag/drop any DAG JSON.
  • Use the toolbar buttons to pull the latest CRISPR, Prime, or PCR demos (generated directly from the simulator). You can also append ?demo=crispr|prime|pcr or ?json=URL to auto-load a DAG.
  • Sliders for probability/time filters.
  • Hover tooltips show stage metadata and root→node diffs.
  • Export PNG or copy a shareable permalink.
  • Rebuild the hosted demos anytime with:
PYTHONPATH=src python3 scripts/generate_playground_dags.py

Side-by-side compare

helix edit-dag compare --a dag_before.json --b dag_after.json --out compare.png \
  --label-a "CRISPR" --label-b "PCR" \
  --min-prob 0.02 --max-time 8 \
  --summary diff.json
  • Unique-to-A branches glow teal, unique-to-B branches glow amber, and shared paths stay neutral.
  • Timeline layout + probability filters keep dense DAGs readable.
  • Inset panel shows the top terminal nodes from each artifact (with normalized probabilities) so you can quantify shifts at a glance.
  • --summary diff.json writes a JSON diff (unique/shared nodes, terminal outcomes, top probabilities) for changelog/PR use.

Perfect for baseline vs redesign comparisons, CRISPR → PCR chaining, or “what-if” pegs.

Dataset generator

helix edit-dag generate-dataset --mechanism crispr --n 25 --out dag_data.jsonl --topk 5
  • Emits JSONL where each row bundles the DAG artifact plus top outcomes/metadata.
  • Mechanisms: crispr (default) or prime.
  • Deterministic via --seed.
  • Handy for training downstream models or stocking the Playground with sample DAGs.

HTML report generator

python -m helix.cli edit-dag report \
  --input dag.json \
  --out report.html \
  --png dag.png

Generates a single HTML artifact with:

  1. Summary metrics (entropy, intended vs unintended probability, node/edge counts).
  2. Embedded PNG/GIF (or fetches from --png).
  3. Top outcome table with diffs and probabilities.
  4. JSON download links for transparency.

You can host the HTML as-is or embed it in docs dashboards.

Artifact schema v1.1 highlights

  • Deterministic node IDs (content-addressed hash of parents + event + stage + time).
  • seq_hashes and diffs per node for compact, tamper-evident storage.
  • Metadata: schema_version, rule_version, created_at.
  • Fully compatible with earlier tools; dag_from_payload reconstructs views lazily.
  1. Run simulation (e.g., helix crispr dag ... --json dag.json).
  2. Render PNG/GIF (helix edit-dag viz/animate).
  3. Generate HTML report (helix edit-dag report).
  4. Share via Playground (host dag.json and link ?json=).

This pipeline yields reproducible, explainable edit twins that teams can inspect, compare, and discuss.