The ascension of AI in R&D
While the AI field has always been forward-looking, its revolution in chemistry and drug discovery has yet to be fully realized. While there have been some high-profile setbacks, several breakthroughs should be watched closely as the field evolves. Generative AI is impacting drug discovery, machine learning is being used more in environmental research, and large language models like ChatGPT are being tested in healthcare applications and clinical settings.
Many scientists are keeping an eye on AlphaFold, DeepMind’s protein structure prediction software that revolutionized how proteins are understood. DeepMind and Isomorphic Labs have recently announced how their latest model shows improved accuracy, can generate predictions for almost all molecules in the Protein Data Bank, and expand coverage to ligands, nucleic acids, and posttranslational modifications. Therapeutic antibody discovery driven by AI is also gaining popularity, and platforms such as the RubrYc Therapeutics antibody discovery engine will help advance research in this area.
Though many look at AI development excitedly, concerns over accurate and accessible training data, fairness and bias, lack of regulatory oversight, impact on academia, scholarly research and publishing, hallucinations in large language models, and even concerns over infodemic threats to public health are being discussed. However, continuous improvement is inevitable with AI, so expect to see many new developments and innovations throughout 2024.