AGENT SKILL

web-ui

Use, deploy, debug, and inspect DiffDock's Gradio web UI and downloadable output archives.

License 许可
MIT
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
web-ui
license
MIT
metadata
{ "disco-role": "operating" }
description
Use, deploy, debug, and inspect DiffDock's Gradio web UI and downloadable output archives.
disable-model-invocation
true

02 资源文件 4

03 SKILL.md 正文

DiffDock Web UI

Use this sub-skill when a user wants to launch DiffDock's Gradio interface, understand the single-complex UI inputs, debug app startup/runtime issues, or inspect the zip returned by the UI.

Start Here

Scope And Routing

This sub-skill covers the Gradio app only: one protein plus one ligand, optional YAML config upload, sample count control, app launch/deployment constraints, and the UI zip/result visualization conventions.

Route these tasks elsewhere:

  • Full command-line inference flags, CSV/batch inference, model checkpoint options, and default_inference_args.yaml tuning: ../docking-inference/SKILL.md.
  • Shared installation, CUDA/Torch/RDKit/PyG/ESM/OpenFold/ProDy dependency failures: ../../references/troubleshooting.md.
  • Training, benchmark evaluation, dataset preparation, or confidence-model development: the relevant sibling sub-skill.

Quick Commands

# Launch the app from a DiffDock checkout after installing the runtime stack.
python app/main.py

# Use a custom Gradio port.
GRADIO_SERVER_PORT=7861 python app/main.py

# Inspect a downloaded DiffDock UI zip.
python sub-skills/web-ui/scripts/inspect_diffdock_output_zip.py diffdock_output.zip

Key Facts

  • The web UI is a convenience wrapper around DiffDock inference for one complex at a time; use command-line inference for batch workflows.
  • Protein input can be a PDB ID or an uploaded .pdb; ligand input can be a SMILES string or an uploaded .sdf/.mol2 file.
  • If a PDB ID is supplied and no protein file is uploaded, the app downloads the PDB from RCSB before running inference.
  • The optional YAML config replaces the default inference config path; it must contain every argument required for the intended run.
  • The sample-count UI field maps to samples_per_complex and is limited to integer values from 1 through 100 in the app.
  • Directly importing the app module from outside the app directory can fail because mol_viewer is imported as a sibling module; launch through the app entrypoint or ensure the app directory is importable.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351