Using 'DNA Movable-Type Printing' to Build Biological Data for AI, Yongyuan Hesheng Secures RMB 50 Million Angel Round | 36Kr Exclusive

Author: Hu Xiangyun Editor: Hai Ruojing 36Kr has learned that Yongyuan Hesheng Technology (Shenzhen) Co., Ltd. ("Yongyuan Hesheng") recently completed an angel round of financing worth RMB 50 million. The round was jointly led by Shunwei Capital and Yunmu Capital. The capital raised will be used to produce high-value datasets, expand production capacity for its self-developed innovative DNA synthesis technology, and further build data-forging infrastructure. Yongyuan Hesheng positions itself as a "data-forging platform for AI and biology," providing research institutions, pharmaceutical companies, and model teams with integrated services spanning sequence construction, functional experiments, and data curation, and delivering AI-ready sequence-function datasets. Founder and CEO Yang Zongying explained that the reason for choosing this direction is that, for AI biological design teams, after generating candidate sequences there is still a question that must be answered through experiments: how these designs actually perform under specified conditions, and which directions should be optimized in the next round? "We hope to take on this step, connecting DNA construction, functional testing, and data curation, and producing sequence-function data around customers' R&D questions. The data can be used both for model training and for evaluating models, comparing candidate solutions, and planning the next round of experiments." Image source: Yongyuan Hesheng In 2015, while pursuing a master's degree in synthetic biology at Imperial College London, Yang Zongying found that although interest in the field was soaring, laboratory research paradigms remained rather primitive. "For example, human cells contain more than 20,000 genes, but researchers often alter only a small number of gene nodes and then observe system changes, so the information they can obtain is very limited." Since relying solely on human "handcrafting" makes it difficult to decode complex biological systems, could AI, which excels at uncovering patterns from massive data, bring a new breakthrough? With this idea in mind, Yang went on to the University of Newcastle in the UK to pursue a PhD in computing, researching genome design using machine learning. Not long after starting the research, he realized that compared with computing power and algorithms, the real gap lay in the large volumes of biological data needed to train models. The scarcity of such data is largely due to a bottleneck at the initial stage: DNA synthesis is too expensive. Mainstream DNA synthesis methods on the market include chemical synthesis, chip-based high-throughput synthesis represented by Twist, and the emerging TdT enzymatic synthesis approach. Each of these pathways has its own technical strengths, but overall, across the three dimensions of cost, speed, and synthesis length, they struggle to fully support the massive data supply AI requires. It was for this reason that Yang