BioMap has reached a NewCo deal worth tens of millions of US dollars with Biogend Therapeutics, with Legend Capital and Temasek Life Sciences Accelerator teaming up to place their bets on the transaction.
Targeting the hundred-billion-dollar autoimmune disease market, Biogend leverages AI to unlock new space for programmable drug discovery
Biogend is an asset-centric AI-driven antibody drug R&D company. Relying on its AI-enabled macromolecule discovery capabilities, it is committed to advancing First-in-Class and Best-in-Class innovations in highly barrier target combinations and novel molecular mechanisms, accelerating the R&D and translation of innovative drugs for the global market.
Autoimmune diseases cover nearly 100 disease types, and the global pharmaceutical market is expected to reach approximately 176 billion U.S. dollars by 2030, while the Chinese market is expected to approach 25 billion U.S. dollars. Evolving from traditional broad-spectrum immunosuppression to targeting inflammatory pathways, and then to more precise restoration of immune balance, the industry is searching for a new generation of treatment solutions that combine deep efficacy, safety and long-term benefits. This also provides a more valuable application scenario for AI-powered pharmaceutical R&D: locking the engineering-accessible optimal solution in an ultra-large combination space, and realizing programmable function regulation through proactive design.
Through cooperation with BioMap, Biogend has accessed BioMap's AI discovery system. This system is underpinned by BioMap's self-developed life science foundation large model xTrimo with 268 billion parameters and the BioMap OS dry-wet closed-loop discovery system, enabling more efficient design, optimization and verification in the candidate molecular space far exceeding the capacity of traditional experimental screening, thus significantly improving the discovery efficiency of macromolecules with programmable and precise functional regulation. This cooperation started at the platform level and was quickly translated into a definable and promotable core asset portfolio. Based on the AI discovery system introduced from BioMap, Biogend has completed the R&D of its first batch of core pipelines and key preclinical verification, made remarkable breakthroughs in the biased modification of the classic cytokine IL-2, and constructed the world's first Treg-biased IL-2 bispecific antibody for autoimmune diseases.
From 10-fold to 390-fold: AI begins to break through the most difficult engineering boundary of IL-2
IL-2 is one of the most classic cytokines in immunology, and also the earliest approved cytokine for pharmaceutical use in human history. It can not only promote the activation of immune cells such as effector T cells and enhance immune response, but also promote the proliferation and activation of Treg cells, participating in the maintenance of immune regulation and tissue tolerance. For this reason, IL-2 is often regarded as a molecule that can both "step on the accelerator" and "step on the brake". The 2025 Nobel Prize in Physiology or Medicine was awarded to three scientists who made key discoveries around peripheral immune tolerance, further highlighting the core value of Treg in immune homeostasis regulation.
For autoimmune diseases, the ideal IL-2 drug does not simply increase activity, but makes the signal more precisely biased towards Treg cells, so as to achieve a more stable and controllable immune regulation effect while minimizing non-selective immune activation. However, the development of IL-2 drugs has long been limited by problems such as narrow therapeutic window, short half-life and high toxicity, making it one of the most challenging engineering topics in immune drug R&D.
Traditional IL-2 optimization mainly relies on site-directed mutation and mutation library screening and verification. A protein usually consists of dozens to thousands of amino acids, and each site has multiple mutation possibilities. The theoretical combination space is extremely huge, far exceeding the scope that traditional experiments can systematically traverse. Although traditional methods can improve receptor selectivity to a certain extent, they can usually only widen the affinity gap between different receptor subunits by about 10 times, and it has always been difficult to substantially break through the capability boundary of selective signal regulation.
Wei Yuan, Head of Molecular Discovery Business at BioMap, introduced that the BioMap team judged at the early stage of the project that IL-2 is a very suitable molecule for AI modification, because the problem definition of this project is clear enough, which is essentially directional optimization around the affinity of different receptor subunits, and can give the system a clear task direction. Focusing on the clear design goal of "enhancing the binding of one type of receptor subunit and weakening the binding of another type of receptor subunit", the R&D team transformed the originally highly complex molecular optimization problem into an engineering task that can be systematically defined, calculated and iteratively verified.
In this project, using BioMap's AI discovery system, the R&D team started from tens of thousands of in silico simulated potential candidate molecules, and after three rounds of dry-wet closed-loop design and experimental verification, finally screened out IL-2 variant molecules with ultra-high bias. Compared with wild-type IL-2, the lead molecule increases the affinity gap between different receptor subunits to more than 390 times, significantly breaking the optimization upper limit of about 10 times by traditional methods, showing outstanding Best-in-Class potential.
Preclinical data further shows that the lead molecule presents biased immunomodulatory characteristics in line with the design expectations: compared with wild-type IL-2 and similar control molecules, its functional bias towards Treg cells is further enhanced, and it shows significantly better anti-inflammatory effects and safety profiles than the control molecules in multiple in vitro and in vivo animal efficacy models.
This project shows that the role of AI in innovative drug R&D is evolving from improving the efficiency of a single link to the overall optimization of complex design goals. By exploring candidate molecules in a broader sequence space and taking into account multiple constraints such as affinity, selectivity, expression and druggability, the team has realized the systematic design of the "Treg bias" goal, and also provided a replicable R&D idea for more complex protein engineering problems.
Redefine engineering efficiency with AI, and become an "innovative drug engineering merchant of First-in-Class and Best-in-Class"
Based on this financing and cooperation, Biogend will continue to deepen the R&D paradigm of AI-driven drug discovery, quickly screen candidate pipelines with First-in-Class and Best-in-Class potential, and promote the commercialization of high-value assets. Based on the existing pipeline layout, Biogend's development direction is gradually becoming clear: focusing on the treatment scenarios of major autoimmune diseases with unmet long-term demands, Biogend will subsequently focus on immune tolerance reconstruction and precise regulation of inflammatory pathways, prioritize the layout of a new generation of antibody pipelines with both clinical transformation potential and global development value, focus on treatment upgrading opportunities for a broad patient population, and promote the R&D goal from simply relieving symptoms to deeper inflammatory control and long-term benefits.
Relying on the underlying capability support provided by BioMap's AI discovery system at the early R&D stage, as well as the support from Legend Capital and Temasek Life Sciences Accelerator in industrial collaboration, resource linkage and international expansion, Biogend is committed to deeply integrating AI and biology, continuously transforming complex immunological problems into calculable, verifiable and engineerable innovative drug pipelines, bringing safer and more durable treatment options for patients with autoimmune diseases around the world.
Liu Wei, Co-founder and CEO of BioMap, said: "BioMap has always focused on the construction of life science foundation large models and AI discovery systems, hoping to serve different types of pharmaceutical companies and biotechs with its core technology base to help partners accelerate innovative discovery. Biogend has a clear pipeline strategy in the field of autoimmune diseases. This cooperation embeds the xTrimo life science foundation large model and BioMap OS dry-wet closed-loop discovery system into Biogend's R&D system, accelerating the continuous accumulation and value output of its pipeline assets. We believe that BioMap's AI discovery system can not only empower single molecule discovery, but also enable biotechs like Biogend that drive pipeline development with AI to build truly differentiated innovation capabilities. We look forward to exploring a transformation path from technical capabilities to innovative assets together with Biogend, so that AI can truly serve high-quality drug R&D."
Hu Haoyue, CEO of Biogend, said: "The immunomodulatory mechanism of autoimmune diseases is complex, clinical demands vary significantly, and there is still a large amount of treatment space to be broken through. Biogend is positioned as an AI-driven, asset-oriented innovative biotechnology company: starting from unmet clinical needs, it quickly obtains innovative molecular discovery capabilities with the help of leading external technology platforms, while concentrating its own resources on disease insight, asset strategy and clinical transformation. The cooperation with BioMap is a key step in the construction of Biogend's innovative R&D system. The positive progress of the Treg-biased IL-2 bispecific antibody project also proves the phased success of this R&D model in terms of efficiency and innovation. In the future, Biogend will continue to expand its differentiated pipeline portfolio, and promote the development and commercialization of high-value pipelines through global cooperation to benefit global autoimmune patients at an early date."
Hong Tan, Managing Director of Legend Capital, said: "Biogend is a key strategic layout of Legend Capital in the tracks of AI macromolecule innovative drugs and autoimmune diseases. This cooperation not only technically confirms the innovation of AI pharmaceuticals on the traditional R&D paradigm, but also realizes the multi-pipeline comprehensive cooperation at the AI platform level in terms of model. Legend Capital has long been firmly optimistic about the long-term dividend of the integration of AI and biomedicine, and continues to bet on innovative biotechnology enterprises with underlying technical barriers and clinical transformation capabilities. In the future, we will continue to empower Biogend, link BioMap's AI technology platform, accelerate the promotion of multiple autoimmune innovative pipelines to clinical verification, and build an AI-driven autoimmune antibody R&D enterprise."
Peter Chia, Chief Executive Officer of Temasek Life Sciences Accelerator, said: "Autoimmune diseases are still one of the most complex cutting-edge fields in modern medicine. Breakthroughs in this field are inseparable from rigorous scientific exploration, as well as the courage to innovate traditional drug R&D paradigms. Biogend has always adhered to a professional development path focused on core assets, and its interleukin-2 (IL-2) project has also obtained preliminary scientific verification, which is very encouraging for us. As a subsequent investor, Temasek Life Sciences Accelerator is pleased to support Biogend to further expand its R&D pipelines, and looks forward to working with the company to transform the potential of the technology platform into innovative therapies that benefit patients around the world, and effectively improve the treatment effect of patients."