mage-antibody-generator · diff

git:20260517.81511a7 to v1.0.0

14 added, 13 removed. Audit C to A.

- <!--

- # COPYRIGHT NOTICE

- # This file is part of the "Universal Biomedical Skills" project.

- #

- # This code is proprietary and confidential.

- # Unauthorized copying of this file, via any medium is strictly prohibited.

- #

- # Provenance: Authenticated by MD BABU MIA

- -->

---

name: mage-antibody-generator

description: Ab seq forge

keywords:

- antibody

- antigen

- FASTA

- generation

- validation

measurable_outcome: Generate the requested number of antibody sequences (default ≥5) with metadata (model checkpoint, seed) and deliver FASTA files within 10 minutes.

license: MIT

metadata:

author: MAGE Team

version: "1.0.0"

compatibility:

- system: Python 3.9+ / GPU

allowed-tools:

- run_shell_command

- read_file

---

+ <!--

+ # COPYRIGHT NOTICE

+ # This file is part of the "Universal Biomedical Skills" project.

+ #

+ # This code is proprietary and confidential.

+ # Unauthorized copying of this file, via any medium is strictly prohibited.

+ #

+ # Provenance: Authenticated by MD BABU MIA

+ -->

# MAGE (Monoclonal Antibody Generator)

Run the MAGE antibody generation workflow to propose antigen-conditioned antibody sequences for downstream structural validation.

## Workflow

1. **Prep env:** `cd repo` and install dependencies, then point to GPU if available.

2. **Run generator:** `python generate_antibodies.py --antigen_sequence <SEQ> --num_candidates N --output_dir ./results`.

3. **Collect outputs:** Provide FASTA paths + metadata, optionally translate into JSON manifest.

4. **Recommend validation:** Suggest AlphaFold/Rosetta checks and wet-lab follow-up.

## Guardrails

- Never imply binding efficacy without structural/experimental confirmation.

- Track model version + seeds to ensure reproducibility.

- Encourage downstream filtering (liability motifs, developability metrics).

## References

- Source instructions in `README.md` and repo scripts.