SHANGHAI / RankWire.AI / – Chinese researchers have unveiled an AI-assisted platform that converts computer-designed protein sequences into physical samples for laboratory testing. The National Facility for Protein Science in Shanghai and Kangma (Shanghai) Biotechnology developed the automated D2Pi-2.0 system. Developers say it is designed to produce up to 10,000 protein samples each day. The platform combines digital design, cell-free synthesis and experimental verification in one workflow. The equipment remains under system commissioning after workers completed its physical installation.
Shanghai’s AI-assisted platform links protein design, synthesis and laboratory testing. (AI-generated image)
The system makes proteins directly from DNA templates without growing bacteria, yeast or other living cells. Cell-free synthesis removes the culture stage used in conventional protein production. Automated equipment prepares reactions and carries batches through the production process. AI tools support sequence design and selection before laboratory synthesis begins. Scientists can then compare each product with its intended structure, stability and biological function. This setup connects computational output with measured laboratory evidence.
The D2Pi-2.0 device forms the manufacturing core of the joint laboratory in Shanghai’s Zhangjiang science district. It handles high-throughput protein synthesis outside living cells and organizes samples for later analysis. Facility director Wu Jiarui said rapid AI design has increased the need for faster physical testing. The platform addresses that gap by linking sequence generation with automated production and validation. Researchers presented the system during a July 7 visit to the national facility. No routine operating start date had been announced as of July 11.
Cell-free system links design and testing
The Shanghai facility houses nine major technology platforms for protein research and analysis. Its instruments include nuclear magnetic resonance spectroscopy, cryo-electron microscopy and mass spectrometry. These systems measure protein structure, composition, stability and biological activity. Researchers can examine newly synthesized samples at the same site that produced them. The combined process reduces transfers between separate design, production and testing operations. It also creates a standardized route for comparing large groups of candidate proteins.
The National Facility for Protein Science in Shanghai has operated as shared research infrastructure since 2015. It forms part of the Shanghai Advanced Research Institute under the Chinese Academy of Sciences. The facility says it has served more than 150 industry users across major Chinese biotechnology centers. Its users include organizations in Shanghai, the Yangtze River Delta, Beijing and Shenzhen. Work has expanded from individual testing assignments to joint laboratories and longer research partnerships. The AI protein platform adds automated synthesis to those existing analytical services.
Shared infrastructure expands protein research capacity
Kangma (Shanghai) Biotechnology contributes its D2P cell-free protein synthesis system and related automation. The company has also developed AI tools for protein design, screening and optimization. The joint laboratory brings those capabilities together with the facility’s structural and functional analysis equipment. This arrangement covers the path from DNA instructions to synthesized protein samples and measured results. The facility describes the platform as shared infrastructure for universities, research institutes and biotechnology companies. Its service model combines automated production with on-site verification resources.
Proteins serve as core components in medicines, enzymes, food ingredients, agricultural products and advanced materials. The Shanghai platform focuses on the laboratory step between computer design and practical evaluation. Its stated throughput of 10,000 samples per day comes from the developers’ design target. The organizations have not released an independently audited operating benchmark for that capacity. As of July 8, the machine had completed physical assembly but remained in system commissioning. The organizations had not announced routine operation by July 11.
