AI and the future of pharmaceutical research (2107.03896v1)

Abstract: This paper examines how pharmaceutical Artificial Intelligence advancements may affect the development of new drugs in the coming years. The question was answered by reviewing a rich body of source material, including industry literature, research journals, AI studies, market reports, market projections, discussion papers, press releases, and organizations' websites. The paper argues that continued innovation in pharmaceutical AI will enable rapid development of safe and effective therapies for previously untreatable diseases. A series of major points support this conclusion: The pharmaceutical industry is in a significant productivity crisis today, and AI-enabled research methods can be directly applied to reduce the time and cost of drug discovery projects. The industry already reported results such as a 10-fold reduction in drug molecule discovery times. Numerous AI alliances between industry, governments, and academia enabled utilizing proprietary data and led to outcomes such as the largest molecule toxicity database to date or more than 200 drug safety predictive models. The momentum was recently increased by the involvement of tech giants combined with record rounds of funding. The long-term effects will range from safer and more effective therapies, through the diminished role of pharmaceutical patents, to large-scale collaboration and new business strategies oriented around currently untreatable diseases. The paper notes that while many reviewed resources seem to have overly optimistic future expectations, even a fraction of these developments would alleviate the productivity crisis. Finally, the paper concludes that the focus on pharmaceutical AI put the industry on a trajectory towards another significant disruption: open data sharing and collaboration.