Artificial intelligence empowered biological research: Paradigm shifts, application scenarios, and strategic layout

Life related processes are characterized by high dimensionality and multi-scale properties. Understanding mechanisms underpinning life processes helps promote national healthcare, agricultural development, sustainable ecological civilization, and national security. Current life science research is confronted with an enormous challenge of dimensionality stemming from data explosion and data fragmentation, for which the recent rapid advancement of artificial intelligence (AI) provides novel solutions. AI will catalyze a paradigm shift in life science research from the current experiment based empirical induction to a new closed-loop knowledge acquisition including large scale data collection, model building, model prediction, experimental validation, and iterative of these procedures. Life science research is moving beyond fragmented descriptive exploration toward a systematic, predictable, and design-driven era. The major biological research fields that AI can provide immediate push include digital cell, health management, diagnosis and treatment, pathogen detection, crop breeding, brain-computer interfaces, ecological management, etc. To secure a competitive edge in global life science research, government now needs to develop national-level AI-friendly data warehouse and computing infrastructure, life science foundational models, as well as agent-based research platforms for both wet-lab and model-based studies. In addition, the interdisciplinary collaboration, and strong bioethics and safety governance system need to be strengthened.