Katrina Choe, Co-Founder and CEO of AI drug discovery firm Syntheia, recently highlighted the significant hurdles in translating artificial intelligence advancements into approved pharmaceutical products. During an Axios Pro: Biotech interview with reporter Andrew Dunn, Choe expressed uncertainty regarding the ultimate clinical success of AI-generated drug candidates, despite the technology's impressive capabilities in early-stage research. Her remarks underscore a growing industry sentiment about the gap between preclinical promise and patient-ready treatments.
Choe acknowledged AI's powerful contributions to drug development, stating: > "We built a lot of things... Yes, we are able to improve the binding of the antibodies. Sure, yes we can do that. We can find the small molecules that are probably going to be better drugs." Syntheia, under Choe's leadership, leverages AI and machine learning to identify novel drug candidates and optimize existing ones, aiming to reduce the time and cost associated with bringing new therapies to market.
However, the CEO immediately tempered this enthusiasm with a dose of reality regarding clinical translation. > "But when will any one of those, or if any one of those, will become a drug? I don’t know. It’s very difficult to tell," Choe added, emphasizing the unpredictable nature of human trials. This sentiment reflects the broader challenges facing the pharmaceutical industry, where the success rate of drug candidates entering Phase 1 trials and eventually gaining approval is notoriously low, often falling below 10%.
Industry experts concur that while AI has accelerated early-stage processes like target identification and lead optimization, it has not yet dramatically improved overall success rates in clinical development. The inherent biological complexities, unpredictability of human physiology, and challenges related to toxicity and pharmacokinetics continue to pose significant barriers. Despite substantial investment in AI, a clear breakthrough in clinical trial success rates directly attributable to AI remains to be widely observed.
Choe's comments from the Axios Pro: Biotech interview serve as a crucial reminder that while AI is revolutionizing the initial phases of drug discovery, the journey from a promising AI-identified compound to a safe and effective medicine for patients remains a complex and arduous endeavor. The industry continues to seek more robust validation mechanisms to bridge this critical gap.
