OMGenomics Podcast 011: AI in biology: distinguishing hype from reality
Aug 13, 2025
Valerie de Crissy-Lagard, a seasoned scientist specializing in enzyme function prediction, and Rachel Thomas, co-founder of Fast AI and an AI expert pursuing a PhD in immunology, discuss the promising yet precarious role of AI in biological research. They delve into a case where AI's enzyme predictions fell short and emphasize the dire need for collaboration between machine learning experts and biologists. With insights on automation bias and the intricacies of genomic annotation, they highlight the importance of a multidisciplinary approach to enhance research quality.
AI Snips
Chapters
Transcript
Episode notes
Long Hunt For A Mischaracterized Protein
- Valerie de Crécy-Lagard hunted the YCIO paralog's function for about 15 years before discovering a paper claiming AI solved it.
- She immediately recognized the paper's claim as the same common mistake students make about paralogs.
In Vitro Activity Is Not Definitive
- In vitro activity alone can mislead because paralogs retain residual ancestral activity under nonphysiological conditions.
- Biological function needs contextual evidence beyond tube assays to be convincing.
Escalate And Publish A Thorough Rebuttal
- If you find a suspect AI claim, follow journal rebuttal channels: contact authors, then editors, and prepare a formal response if needed.
- When reviewers ignore specialist errors, publish a thorough corrective study with curated analyses.