Shares of METIS TECHBIO-P (07666) climbed more than 7% during intraday trading. At the time of writing, the stock was up 3.85%, trading at HK$17.53, with turnover reaching HK$81.72 million.

On September 19, the 29th Beijing-Taiwan Science and Technology Forum was held in the Beijing Economic-Technological Development Area. METIS TECHBIO-P co-founder, Chairman, and CEO Lai Caida was invited to deliver a keynote speech titled "The Era of Programmable Biology: AI's New Frontier." Rather than focusing on "how large the model is," Dr. Lai emphasized another critical point: the ultimate competition in biomedical AI hinges on the ability to continuously obtain high-quality experimental data and transform that data into model capabilities.

It is understood that Dr. Lai described the technical architecture of AI-driven drug development using a "three-layer cake" analogy. The bottom—and thickest—layer is the data platform for dry and wet experiments. While general-purpose large models can access vast amounts of text from the internet, the data truly needed for drug delivery, protein, nucleic acid, and cell research largely originates in the laboratory: which tissues an LNP targets, how different lipid structures affect expression, and how protein sequence changes alter activity—all of these must be validated experimentally.

The second layer consists of vertical domain models generated from this data. The third layer comprises large language models and agents. Since launching NanoForge, the world's first AI-powered nano-delivery platform, METIS TECHBIO-P's pioneering "dry experiment + wet experiment + intelligent agent" R&D paradigm has comprehensively covered the complete chain of AI-driven protein generation, in vivo delivery technology, and next-generation immunotherapy development. This forms a self-learning R&D loop and a nano-delivery data moat, achieving continuous breakthroughs in extrahepatic delivery and providing foundational support for AI-precise delivery and cellular programming exploration.