How AI is changing antibiotic discovery
Anchor workflow: César de la Fuente's lab uses computational and AI tools to search living and extinct genomes for antimicrobial candidates. Candidate generation is not the same as a validated medicine.
Primary / high-quality sources: OpenAI · de la Fuente lab · NIH · ApexGo research context
Antimicrobial resistance creates continuous pressure for new defenses. Biology contains enormous molecular diversity, but experimentally testing every possibility is impossible.
Genomes can be searched computationally for sequences with properties associated with useful peptides. AI can prioritize candidates hidden in data that would be impractical to inspect manually.
Extinct-organism sequences expand the searchable library beyond molecules present in living species. The idea is not to recreate extinct organisms, but to inspect ancient biological information for potentially useful molecules.
AI can both identify candidates and propose modifications intended to improve activity. These are computational hypotheses until laboratory experiments test them.
A molecule that looks promising computationally must still survive synthesis, activity testing, toxicity studies, pharmacology, animal studies, manufacturing, and eventually clinical trials.
The emerging architecture is Generate → Prioritize → Test → Learn → Redesign. Laboratory results can feed the next computational cycle, progressively focusing the search.
The meaningful metric is not how many molecules AI proposes. It is validated hit rate, improvement over baselines, safety, reproducibility, time saved, and ultimately clinical utility.
Programmable medicine may emerge from coupling machine-scale biological search with increasingly automated experimental validation—not from AI prediction alone.
Hype check: capability is not the same as reliability, a candidate is not the same as a validated outcome, and an observed case is not the same as ecosystem-wide prevalence. Track evidence, replication, permissions and real-world validation.
Enter your name and participate in at least 12 of the 24 practice cases. Then generate the certificate manually.
0 / 24 · 0%
This certifies that
has actively participated in
Juan Carballo
Training Content Curator & Interactive Learning Designer