"Microelement Synthesis" Secures 300 Million Yuan in Series A+ Round of Financing, Jointly Launches an Open and Collaborative Platform for AI Biocomputing | First Report by 36Kr

36Kr learned that in March 2026, "Weiyuan Synthesis" completed a 300 million yuan Series A+ financing round, with investment from Henan Investment Group Huirong Fund and Mr. Tan Ruiqing. This round of financing will help Weiyuan Synthesis expand the application boundaries of AI biological computing, increase investment in core technology R & D, and promote scenario-based implementation.

It is reported that previously, Henan Investment Group had completed an in - depth layout of the AI infrastructure industry. From investing in chips, controlling ultra - large - scale computing power, to comprehensively integrating HALO assets, it has provided power and computing support for the implementation of AI scenario applications. It is worth noting that the investor in this round, Tan Ruiqing, is a board member of Fudan University, the president of the alumni association of the Department of Chemistry, the co - founder and former vice - chairman of the well - known A - share listed company Longbai Group, and has been deeply involved in entrepreneurship and investment in the chemical industry for more than 30 years.

As the life sciences fully enter the AI4S era, breakthroughs in underlying technologies such as protein structure prediction and design are changing the R & D paradigm of the biomanufacturing industry. Currently, applying AI algorithms and model capabilities to the real world of biomanufacturing, especially in scenarios such as enzyme engineering and metabolic pathway optimization, has become an important focus of the industry.

In January 2026, Weiyuan Synthesis, in collaboration with institutions such as Stanford University, Princeton University, Peking University, and ByteDance, published the latest results at the top - level academic conference on artificial intelligence, ICLR 2026: an open - collaboration platform for global scientists, PoseX (platform address: http://dock - lab.tech/), aiming to solve the problem of molecular docking in real - world scenarios and provide a fair and real assessment of the capabilities of different docking algorithms and models.

In synthetic biology and drug R & D, molecular docking is one of the underlying core technologies. Accurately predicting the binding mode of ligands and receptor proteins is like finding a key to open the life factory in the microscopic world. In the past, this work highly relied on scientists' experience or costly physical simulations.

In an ideal experimental environment, putting the ligand from a known co - crystal structure back into its original pocket is like filling in a puzzle. However, in the real battlefield of enzyme engineering design, proteins are not static "locks" but "jellies" that constantly change shape. Side chains rotate, backbones breathe, and pocket shapes may be completely reshaped. How to accurately predict the binding mode under the dynamic changes of protein structures is a recognized challenge.

For a long time, the industrial community has lacked a unified and high - quality benchmark to evaluate the performance of algorithms in cross - conformation scenarios. To define the "actual combat standard", the PoseX platform has built a large - scale, open - source docking evaluation platform that is close to real R & D scenarios to solve problems such as "single benchmark data, poor generalization, and deviation from actual application scenarios".

At the same time, PoseX tested 24 mainstream methods, including physical methods such as Schrödinger Glide, AI docking methods such as DiffDock, and AI co - folding methods such as AlphaFold3 and Chai.

After testing, the joint R & D team concluded that "top - notch AI docking methods (such as SurfDock) and co - folding methods (such as AlphaFold3) have comprehensively surpassed the physical models that have dominated the industry for decades in terms of accuracy and robustness when dealing with the most challenging cross - conformation docking tasks."

In August 2025, Liu Bo, the founder of Weiyuan Synthesis, told 36Kr that the mechanisms of enzyme design and metabolic network optimization are very complex, and it is difficult to have a "unified" algorithm model to solve all problems. Therefore, it is necessary to select the most suitable model to accelerate the R & D of specific links according to specific project scenarios. At that time, he had built an AI R & D team of about 15 - 20 people and cooperated with top - level global AI algorithm laboratories to conduct benchmarks using their wet and dry experimental capabilities.

In his view, with the opening of algorithms and the relative reduction of computing thresholds, the importance of wet experiments is particularly prominent. "How to evaluate different enzymes, how to select expression systems, how to set strict test conditions, and verify them with high - throughput equipment, these involve high barriers, and we have built a complete system."

In the process of jointly launching the PoseX platform, Weiyuan Synthesis continuously tests and feeds back real biological experimental data to top - level global model teams. While establishing a standardized evaluation platform for "protein - ligand docking", it also applies the proven tools to its own pipeline development.

With the help of the PoseX platform, AI technology is substantially accelerating the R & D and implementation of pipelines from three dimensions: First, accurately simulate the cross - conformation changes of proteins and efficiently develop "super - enzymes" with high temperature resistance and high conversion in the digital space;

Second, combine pocket information and pose refinement to accurately locate and open up key nodes in the optimal metabolic network, realize the metabolic reconstruction of chassis cells and remove bottlenecks, and promote the improvement of production, purity, and cost indicators;

Third, through AI simulation + physical post - processing, compress the time - consuming and costly wet - experiment iteration from months to weeks, improve the return on R & D investment, and reduce the risk of trial and error.

In terms of specific pipelines, Weiyuan Synthesis has currently completed the R & D and industrialization of a number of human and animal nutrition products, such as allulose, lutein, and mannitol. In the field of methanol biomanufacturing, it has developed multiple strains with high - efficiency methanol assimilation and is accelerating the layout of product pipelines including bulk amino acids and biobased material monomers.