GPT-5 Pro Helps Derya Unutmaz Crack 3-Year T-Cell Mystery
Updated
Updated · OpenAI · Jun 23
GPT-5 Pro Helps Derya Unutmaz Crack 3-Year T-Cell Mystery
2 articles · Updated · OpenAI · Jun 23
Summary
Late 2025 use of GPT-5 Pro helped immunologist Derya Unutmaz explain a 2022 T-cell experiment his lab had shelved after failing to interpret why deoxyglucose drove far more inflammatory Th17 cells than low glucose did.
The model pointed to deoxyglucose disrupting production of IL-2, a protein that normally restrains Th17 specialization, suggesting the effect was not just energy deprivation but removal of a molecular brake.
Unutmaz said GPT-5 Pro also correctly predicted the outcome of an unpublished lymphoma experiment, forecasting stronger cancer-killing activity in CD8+ T cells before the result appeared online.
He now uses AI to scan large volumes of research, simulate experiments and narrow hypotheses, arguing it can cut weeks to years from biology work while still requiring expert human judgment.
The findings highlight potential gains for cancer, autoimmune and infection research, alongside OpenAI's warning that faster biological discovery also raises misuse risks that require safeguards.
As AI begins predicting experimental results, how will science guard against the decline of hands-on discovery and algorithmic bias?
AI has solved a complex immunology puzzle. What long-unsolvable biological mystery will be the next to fall?
When an AI generates a novel scientific breakthrough, who gets the credit: the scientist, the AI, or its creators?
From T-Cell Enigma to AI-Driven Insight: The Transformative Impact of GPT-5 Pro on Biomedical Research
Overview
In late 2025, a major breakthrough in immunology was achieved when Dr. Derya Unutmaz and his team, after years of puzzling over the unexpected increase of inflammatory Th17 cells despite deoxyglucose treatment, finally solved the mystery with the help of GPT-5 Pro. The AI analyzed complex biological data and revealed that deoxyglucose was disrupting the construction of the IL-2 protein, which is crucial for regulating T-cells. This discovery not only explained the persistent anomaly but also showcased the powerful potential of human-AI collaboration in advancing scientific research.