BioGeometry Completes Pre-A Round, Applying Generative AI to Bio-Manufacturing

BioGeometry, a frontier digital biology company, recently announced the completion of a new funding round led by The Jiangmen, with Zhipu AI and Shengjing Jiacheng as co-investors. Existing investor Gaorong Ventures also continued to increase its stake. Gaorong Ventures had previously led BioGeometry's angel round in 2022 as the sole investor.

The proceeds from this round will primarily be used to accelerate the deployment of generative AI large models in bio-manufacturing and to advance the development of proprietary products. BioGeometry's long-term vision is to build a generative AI platform that accelerates China's transition toward digitalized, intelligent, green, and sustainable bio-manufacturing.

Bio-Manufacturing Enters a New Era of Digitalization and Intelligence

Bio-manufacturing is increasingly viewed as a key driver of the "Fourth Industrial Revolution," with its market size projected to reach tens of trillions of dollars. The field spans multiple industries including biomedicine, food, chemicals, medical aesthetics, and materials. Its core challenge lies in discovering high-value molecules (such as antibodies, vaccines, and peptide molecules) or optimizing existing microbial strains and biocatalytic enzymes to improve production efficiency and reduce costs. Therefore, the rapid and accurate discovery of high-activity protein macromolecules has become critical across multiple domains of bio-manufacturing.

Advances in artificial intelligence, particularly the emergence of large-scale AI models, are ushering bio-manufacturing into a new era of digitalization and intelligence. This transformation is enabled by cutting-edge biotechnologies such as high-throughput gene synthesis, sequencing, and editing, which are driving exponential growth in biological data and opening vast new possibilities for AI applications.

In the life sciences, following the revolutionary breakthrough of Google DeepMind's AlphaFold2 in protein structure prediction, this year's AlphaFold3 technology — based on diffusion generative models — has extended prediction capabilities to the structures of all biomolecules and their interactions, drawing widespread industry attention.

Building a Generative AI Protein Design Platform

BioGeometry focuses on building a generative AI protein design platform to empower the entire bio-manufacturing sector. Founded in 2022 by Jian Tang, a leading scientist in AI-driven drug discovery, the company counts Yoshua Bengio — known as the "Godfather of AI" and a Turing Award laureate — as its chief scientific advisor. As pioneers in applying generative AI to molecular generation, BioGeometry's core team was among the earliest to use diffusion models for 3D molecular structure generation as early as 2021. Their key paper, GeoDiff, ranked among the top 50 most-cited AI papers in 2022.

Recently, BioGeometry released its latest generative AI protein design large model, GeoFlow, which can simultaneously handle two critical tasks: antigen-antibody complex structure prediction and antibody design. Unlike existing Transformer architectures, GeoFlow employs a geometric deep learning foundation model that better captures atom-to-atom relationships in three-dimensional space. For its generative model, GeoFlow adopts the latest flow matching model, which offers more efficient training and inference compared to diffusion generative models, as well as greater robustness. In antigen-antibody complex structure prediction, GeoFlow has achieved performance on par with AlphaFold3, making it the industry's first generative AI macromolecular design model to approach AlphaFold3-level capability.

GeoFlow Model Architecture Diagram

Antigen-Antibody Complex Prediction Benchmark Results: GeoFlow approaches AF3 performance, with accuracy double that of AF2

Based on its self-developed generative AI large model, BioGeometry has developed GeoBiologics, an end-to-end protein design platform that has already secured licensing partnerships with multiple domestic and international pharmaceutical companies. Additionally, BioGeometry has established strategic collaborations with various upstream and downstream partners to jointly advance the deployment of generative AI in bio-manufacturing, including building AI-powered bio-manufacturing intelligent solutions with Alibaba Cloud, co-developing antibody design generative AI large models with companies such as Sanyou Biopharmaceuticals and Baijun Biotech, and advancing joint development of protein and antibody drug products with Xiangyao Biotech and YinJia Biotech.

GeoBiologics Protein Design Platform

Going forward, BioGeometry will continue to optimize its generative AI protein design platform to empower the bio-manufacturing industry and accelerate the commercialization of proprietary products.

Dr. Jian Tang, CEO of BioGeometry, stated: "We are grateful for the trust from The Jiangmen, Zhipu AI, Shengjing Jiacheng, and our existing investor Gaorong Ventures. Over the past year and a half, BioGeometry has established both a protein generative AI large model platform and a high-throughput protein wet-lab validation platform, with successful validation across multiple projects in antibody, peptide, vaccine, and enzyme design. We will continue to refine our platform and accelerate the launch of our own products."

For more information, visit: www.biogeom.com