This repo is for research target selection and computational experiments.
For medicine / biology:
- Do not present outputs as medical advice.
- Do not recommend treatments to patients.
- Do not run clinical claims without expert review.
- Keep first artifacts to prediction, ranking, literature synthesis, and preclinical assay design.
- Any wet-lab or clinical translation needs domain expert ownership.
The bar is: can we produce a checkable hypothesis, not can we bypass medicine.