How a researcher uses Codex and ChatGPT to search for new antimicrobial molecules

AI summary

Researchers are utilizing deep-learning models to scan vast genome and protein databases for potential antimicrobial candidates, shifting the paradigm from chemical modifications to biological sequence analysis. By treating biological data as an information system, AI accelerates the discovery process from years to hours. The workflow integrates AI tools for sequence pattern recognition, hypothesis generation, and cross-disciplinary data synthesis.

Key points

• Deep learning can identify patterns in biological sequences to predict antimicrobial activity in genome and protein databases.
• Treating DNA and protein structures as information systems allows for the application of linguistic and pattern-matching models to biological discovery.
• AI-driven workflows reduce candidate screening timelines from years to hours by automating data processing and pattern detection.
• Cross-disciplinary breakthroughs are facilitated by using large language models as brainstorming partners to synthesize information across disparate scientific fields.