At the 2026 Frost & Sullivan Summit, the significant session “Synthetic Biology and Biomanufacturing Forum” was organized by Frost & Sullivan and guided by the National Biomanufacturing Industry Innovation Center. The event brought together academicians from two academies, scholars from top universities, leaders of leading enterprises, founders of innovative technology companies, and representatives from prominent investment institutions. They discussed pathways to innovation in synthetic biology, strategies for the industrialization of biomanufacturing, and new opportunities for future growth in the industry. This forum created a high-level and in-depth dialogue platform integrating all sectors—government, industry, academia, research, finance, application, and services—and gathered a strong synergy from basic research to industrial transformation and capital support.
The forum featured a keynote speech by Professor Li Yu from The Chinese University of Hong Kong titled “How Artificial Intelligence and Virtual Cells Can Support Biotech Manufacturing”. Leveraging his interdisciplinary research background and current trends in AI technology, Professor Li Yu analyzed the underlying mechanisms of AI enabling life sciences and synthetic biomanufacturing from perspectives such as general AI development laws, multi-scale life intelligence modeling, and the implementation of virtual cell technologies. He shared the team’s advanced modeling achievements at the levels of biological macromolecules, single cells, and multicellular cells, and proposed an iterative path for the development of synthetic biology DBTL. This provided new technical ideas and practical directions for cross-innovation between AI and synthetic biology, as well as for improving the quality and efficiency of the biomanufacturing industry.
Speaker
Mr. Li Yu
Associate Professor in the Department of Computer Science and Engineering, The Chinese University of Hong Kong
His main research areas focus on bioinformatics, artificial intelligence, and smart health. His research findings have been published in international top journals such as “Nature Biotechnology” and “Nature Methods”, and he has received special reports from “Nature” and “Science”. He graduated from the Bei Shizhang Life Science Talent Class at the School of Life Sciences, Chinese Academy of Sciences, with a first-class honor degree in 2015. In 2016 and 2020, he obtained a master’s and a doctorate from King Abdullah University of Science and Technology (KAUST). In 2022, he was included in Forbes Asia’s 30 Under 30 list; in 2024, he received the President’s Model Teaching Award from The Chinese University of Hong Kong, and was selected as one of the 35 young innovators in China by MIT Technology Review under 35 years old and as an APEC Young Leader under 30 years old. In 2025, he was named a ZIYuan Scholar. He has long focused on machine learning, deep learning, and optimization algorithm development, conducting cross-research on innovative drug development, biological sequence analysis, and element design, and actively promoting the application of AI algorithms in synthetic biomanufacturing.
Key points from Professor Li Yu’s speech:
PART.01
Technical Evolution: General AI Rapidly Evolves, Opening New Opportunities for Intelligent Research in Life Sciences
Professor Li Yu said that focusing on artificial intelligence and virtual cells to enhance synthetic biomanufacturing is an important exploration based on current technological trends. Currently, general AI technology has achieved significant development, completely changing the capabilities of traditional AI. The core criterion used to measure the intelligence level of AI—the Turing test—has now been fully surpassed by large language models, allowing various AI tools to interact naturally like humans.
In addition to language interaction capabilities, AI entity technologies have also made disruptive breakthroughs in recent years. Humanoid robots can closely mimic human walking and climbing behaviors, and their appearance and movements are increasingly similar to those of humans. The speed of technological evolution far exceeds industry expectations. However, Professor Li Yu emphasized that even if AI continues to improve its ability to simulate human behavior and form, there are fundamental differences between its underlying structure and living organisms. AI operates based on metal, hardware, and programs, while the human life system is composed of cells and biological macromolecules, with extremely complex internal mechanisms.
From the perspective of life science research, processes such as DNA transcription, translation, and macromolecule operation within cells seem straightforward, but they are actually highly complex. Traditional life research relies on experimental techniques such as sequencing and imaging to explore life laws, but the integration of AI and life sciences is not simply applying technology. It involves using the hidden laws and patterns of biological systems to train intelligent models through large amounts of biological observation data, precisely identifying the internal logic of molecules, cells, and life systems, providing new intelligent research tools for pharmaceutical development and biomanufacturing.
Professor Li Yu compared the success of natural language large models to point out that human language has fixed grammar and rules, which support the training and implementation of large models. Similarly, biological molecules and cell systems also have stable internal laws. The accumulation of large amounts of biological experimental data provides a solid foundation for AI modeling, discovery of life laws, optimization of biological processes, and design of innovative functions.
PART.02
Deep Research: Building Multi-Scale AI Models, Creating a “Virtual Cell” System for Life Research
Based on the hierarchical characteristics of biological systems, Professor Li Yu’s team has developed AI modeling systems in three dimensions: molecular, single-cell, and multicellular. They have independently developed various intelligent algorithms, gradually establishing a complete “virtual cell” research framework to achieve intelligent analysis of life structures, functions, and dynamic changes.
At the molecular level, the industry’s breakthrough with AlphaFold protein structure prediction demonstrates the great value of AI in biological macromolecule research. Professor Li Yu’s team focused on industry gaps and developed RNA structure prediction models that outperform existing mainstream algorithms, filling the technical gap in intelligent modeling of RNA macromolecules. They also focused on sequence analysis, structural interpretation, and intermolecular interactions of biological macromolecules such as DNA and proteins, improving the intelligent research system for biological macromolecules.
At the single-cell level, the team addressed the challenges of gene expression regulation and cell phenotype prediction. Professor Li Yu explained that all cells in the human body develop from homologous stem cells, and the differences in cell function and morphology stem from differential gene expression. Most diseases and cell function variations are also related to abnormal gene expression. To address this, the team developed specialized AI algorithms that can accurately predict gene expression characteristics of different cell types, simulate dynamic changes in cell gene expression under drugs and genetic manipulation, and predict variations in cell morphology and phenotype. These results have been published in the international top journal “Nature” and provide precise intelligent tools for drug screening and cell function modification.
At the multicellular and tissue level, the academic community is still in the exploratory stage, and the technical difficulty is very high. Professor Li Yu’s team has collaborated with top institutions such as Stanford University to focus on embryonic development systems, establishing an AI modeling framework for multicellular systems to predict developmental states at different stages of embryos and continuously improve related systems. Using multi-scale modeling capabilities, the team has further developed reverse engineering algorithms that allow designing new functional RNA from scratch, screening antibacterial peptides with performance superior to natural products, and breaking through the performance limitations of natural biomaterials, enabling innovative design of functional biological molecules.
PART.03
Industry Empowerment: Virtual Cells Reconstruct Research Paradigms, Facilitating the Upgrade of Synthetic Biomanufacturing
Professor Li Yu pointed out that the team originally focused on AI algorithm research in the field of pharmaceutical development, but the core technical logic aligns well with synthetic biology, enabling seamless cross-industry empowerment and providing essential support for the innovation of biomanufacturing industry.
In terms of technical system compatibility, the molecular, single-cell, and multicellular multi-scale virtual cell framework established by the team matches the full-level research dimensions of synthetic biology—from molecular networks, single-cell strains, multicellular systems to industrial applications. The core capabilities of virtual cells, such as perturbation simulation, function prediction, and element design, can directly be applied to core scenarios such as strain transformation, metabolic pathway optimization, and innovative design of biological elements in synthetic biology.
In terms of research paradigm innovation, traditional synthetic biology relies on the DBTL (Design-Build-Test-Learn) cycle for research, with rigid processes, long iteration cycles, and high trial-and-error costs. The integration of AI and virtual cell technologies completely redefines the research logic, forming a new LDBT (Learn-Design-Build-Test) reverse research paradigm. With existing biological data and models, AI is used to learn and discover life laws, predicting optimal design solutions, followed by experimental design, construction, and testing, significantly reducing research cycles and trial-and-error costs, and enhancing the efficiency of synthetic biology research and manufacturing.
Finally, Professor Li Yu concluded that the core value of AI and virtual cell technologies lies in using intelligent methods to solve the complex laws of biological systems, advancing research from understanding life to designing and transforming it. As multi-scale AI modeling technologies become more mature, these capabilities will be fully applied to the field of synthetic biology, deeply empowering innovation in biological elements, strain transformation and optimization, and manufacturing processes, fully unleashing the innovation potential of the biomanufacturing industry and injecting new momentum into high-quality and intelligent development of the industry.
