Enhancing Artificial Intelligence for the Identification of Healing Agents for Treatment Development
In a recent panel discussion at The Festival of Genomics and Biodata in Boston, experts from academia, pharma, and prestigious institutions such as Rockefeller University, Merck, Sanofi, Dana-Farber Cancer Institute, The University of Texas MD Anderson Cancer Center, and The Institute of Cancer Research in London, discussed the future of Artificial Intelligence (AI) in drug discovery.
The panel, which was led by Professor Bissan Al-Lazikani from The Institute of Cancer Research in London, highlighted the potential of AI in speeding up key steps in the drug discovery process and bringing new treatments to patients. However, they also emphasised the importance of human expertise and quality data in this process.
Professor Al-Lazikani explained, "AI shows potential in drug target predictions, particularly with the integration of new data types such as multi-omics and proteomics data." She further added, "However, over-reliance on artificial or generated inputs in AI-driven therapeutic target discovery is cautioned against."
The quality of biological models is considered crucial in drug discovery, as it remains a bottleneck regardless of the sophistication of AI models. To combat the risk of AI decay when using synthetic data, efforts should be made to generate meaningful biological data.
AI, when paired with high-quality biological models and human expertise, could play a crucial role in drug discovery. However, it is essential to remember that synthetic datasets can help train AI systems, but they cannot replace real-world biological and clinical information.
Even in contexts such as target identification, where AI has drastically reduced drug discovery timeframes, human judgement can help determine which targets are worth pursuing and which aren't. The quote, "We all talk about the nightmare of 95% of cancer drugs entering clinical trials failing. Well, 100% of those succeeded in mice," highlights a significant challenge in drug discovery: the discrepancy between drug success in current models and the final step in humans.
The panelists also discussed the importance of human oversight in AI-driven research. AI cannot replace the expertise of scientists, and human judgement is essential in guiding decisions in AI-driven research.
A paper published by the speakers demonstrated that proteomics data was far superior to any other individual modality at predicting patient drug response. This underlines the importance of quality data in therapeutic target discovery.
In conclusion, while AI holds great promise for drug discovery, it should be used in conjunction with high-quality biological models and human expertise. Simulations using synthetic data risk misleading scientists rather than accelerating discovery, and over-reliance on artificial or generated inputs in AI-driven therapeutic target discovery is cautioned against.
Read also:
- Bishop expresses that Respect Life Month holds additional significance during the Jubilee Year
- Protesters personally deliver correspondence to Fetterman's workplace
- Delay in CDC's Decision on Hepatitis B Vaccine: Insights into Potential Alterations for Parents
- Specialist in Vision Therapy: Expert in Behavioral and Developmental Optometry
