Alan Turing’s question about whether machines can think has led to advancements in artificial intelligence (AI) that are now revolutionizing biomedical research. By using AI methodologies, scientists are able to analyze and draw insights from large amounts of data, leading to new discoveries in human health and disease. CZI recently held a science technology convening focusing on AI for biomedical research, with industry experts discussing the potential of training AI models to expand the understanding of human biology.
The integration of basic science and AI is seen as essential for the future of research, with AI enabling faster and more robust predictions in areas such as immune cell responses to genetic mutations. Multimodal datasets, which combine sequence and image-based information, are also proving to be valuable for training AI models. Generative AI models are being developed to identify molecules for drug discovery, with the potential to combat antibiotic-resistant pathogens.
CZI’s goal is to create a general-purpose model for virtual cells that can transfer information across datasets, serve multiple queries, and unify data from different modalities. This model will provide a foundational resource for biomedical research, allowing scientists to ask complex questions and test theories about human biology more efficiently than traditional methods. By bridging the gap between biological data and AI advancements, CZI aims to create an open and accessible digital platform for biology that will revolutionize the way researchers study and understand human biology.
Source link
Source link: https://chanzuckerberg.com/blog/2024-ai-sci-tech-convening/