Expert Knowledge Base for Computational Biology: Transforming Top Scholars' Experiences into AI-Learnable Skill Packs

This project was released by BioTender-max on GitHub on May 27, 2026 (original link: https://github.com/BioTender-max/computational-biology-experts). Its core is to transform the wisdom of 13 milestone scholars in computational biology into structured SKILL.md skill packs, addressing the problem of difficulty in systematic inheritance of expert experience and providing reusable knowledge assets for AI systems and researchers.

Expert experience and intuition in scientific research are difficult to inherit systematically. As an interdisciplinary field, computational biology has extremely valuable thinking patterns and research methods from top scholars, but this knowledge is scattered across papers, interviews, and conference speeches, lacking a unified way of organization and dissemination.

The project includes 13 outstanding scholars in computational biology, such as statistical genetics expert Alkes Price, single-cell genomics pioneer Aviv Regev, machine learning theory master Bernhard Schölkopf, and protein design leader David Baker, covering key fields like population genetics, single-cell analysis, and protein structure prediction.

SKILL.md Format

Each skill pack includes: 1. Reasoning framework (systematic thinking method); 2. Heuristic methods (practical rules of thumb); 3. Mental models (frameworks for understanding complex biological systems); 4. Anti-pattern warnings (research pitfalls); 5. Classic quotes (incisive remarks from scholars); 6. AI-generated avatars.

Knowledge Extraction

Expert knowledge is obtained and verified through literature mining (analysis of representative papers), interview organization (integration of public speeches), and community validation (revision based on academic feedback).

Skill pack application scenarios: 1. AI-assisted research (simulating scholars' thinking to provide suggestions); 2. Education and training (helping students understand masters' thinking paths); 3. Interdisciplinary reference (other fields refer to structured methods); 4. Research decision support (querying heuristic methods to gain inspiration).

The project demonstrates a direction for knowledge engineering: transforming human tacit knowledge into machine-understandable explicit knowledge, helping to enhance AI capabilities and knowledge inheritance. The SKILL.md format embodies four principles: structured, verifiable (source traceable), extensible (easy to update), and interoperable (supports AI integration).

Limitations

1. Knowledge subjectivity (refinement involves subjective judgment); 2. Timeliness (requires continuous updates); 3. Copyright and privacy (need to handle interviews and portrait rights); 4. Verification difficulty (effect of heuristic methods is hard to quantify).

Outlook

It provides an example for the digital inheritance of scientific knowledge. We hope more fields will draw on this model, combining large language models and knowledge graphs to make knowledge representation more accurate and practical.