ClickMab by Kome Biotech Validated as World’s First AI Platform for De Novo Epitope-Specific Antibody Design Without Reliance on Known Antibodies

Recently, Clickmab, jointly established by XtalPi Innovation Center and Biocytogen, announced a milestone achievement in the development of the world's first internal data-driven AI platform for de novo design of human antibodies. This platform successfully realized AI-driven de novo antibody design without relying on known antibodies. It took only five days to computationally generate novel initial antibody candidate molecules targeting specified epitopes. Further experimental validation showed that these antibodies possess strong affinity and have the potential to advance as PCCs (Preclinical Candidate Compounds) after optimization. Clickmab’s AI-driven de novo antibody design platform overcomes the limitations of traditional wet-lab experiments, reducing the conventional 4-6 month R&D process to just weeks while achieving greater precision and efficiency. The platform also covers "undruggable" antigen targets traditionally difficult to address, bringing a paradigm shift to antibody drug discovery.

Antibody drugs, due to their unique advantages such as stronger targeting, safety, and long-lasting effects, are being increasingly utilized in the treatment of diseases like cancer, autoimmune disorders, and viral infections, gaining significant attention from pharmaceutical companies. According to Global Market Insights, the global antibody drug market exceeded $186.1 billion in 2021 and is expected to maintain a high growth rate in the coming years, reaching $445.6 billion by 2028.

Current antibody development methods rely on wet-lab approaches driven by mouse immunization, phage display screening, and other techniques. These methods are expensive, time-consuming, and lack clear epitope specificity. Not only do they fail to target specific epitopes on antigens based on biological function and drug efficacy, but they also cannot study low-immunogenicity epitopes that are difficult to generate antibodies for in animal immunization experiments. Moreover, numerous challenges remain in later stages of development, such as humanization and developability, making it difficult to obtain ideal antibodies and greatly limiting the research and development of antibody drugs.

Clickmab leverages the AI algorithm development capabilities of XtalPi's Innovation Center and the high-quality fully human antibody sequences generated by Biocytogen’s unique RenMice platform, a fully human antibody mouse. The deep learning generative model trained on this data can simulate the somatic hypermutation process during the antibody generation in the mouse immune system, producing high-quality antibody sequences with high humanization and high affinity. Building on this foundation, Clickmab has established an efficient AI computational workflow for de novo antibody design. By integrating a series of proprietary models, including antigen-specific antibody selection, epitope prediction, and antibody affinity prediction, along with evolutionary algorithms, the company is able to design from scratch candidate antibody molecules with high affinity and specific binding capabilities targeting designated antigen epitopes. Unlike other reported methods of de novo antibody design, Clickmab’s AI-driven antibody design process does not rely on any known epitope-binding antibodies, achieving a groundbreaking transition from 0 to 1 in AI-based de novo antibody design. This significantly expands the space for druggable antibody discovery and has the potential to support XtalPi’s AI-driven antibody drug discovery business.

In a recent test case, for target epitopes not present in the training set, Clickmab designed 1 million antibodies from scratch through an AI computational process and selected 10 for expression, saving hundreds of thousands to millions of antibody experiments with just 5 days of computation. As measured by multi-concentration BLI technology, two of the antibodies showed affinity KD values of 194 nM and 336 nM, respectively, demonstrating significant potential as starting candidate molecules. After further optimization and in vitro/in vivo experimental validation, they can be advanced toward clinical trials.

Clickmab's AI-based de novo antibody generation platform offers the advantages of speed, high accuracy, and low cost, reducing the antibody discovery process from several months to just a few weeks. It is applicable to epitopes on multi-transmembrane target proteins such as GPCRs, which are known for low immunogenicity and difficult stable expression, bringing "undruggable" targets within reach for development and demonstrating extremely high application value. Additionally, the new antibodies are generated based on fully human common light chain mouse antibody data, ensuring a naturally high degree of humanization and excellent developability, which mitigates subsequent CMC risks from the outset. They also effectively address light-heavy chain mismatch issues, providing significant advantages in multi-specific antibody discovery. This achievement validates the feasibility and superiority of AI-generated biologics and lays the foundation for industrial applications of AI-designed functional antibodies.

Clickmab's AI Antibody De Novo Design Platform Success Marks a Boom in AI Antibody Drug Development, First Validating the Innovative Potential of AI Algorithms to Surpass Wet Lab Experiments and Known Sample Library Limitations, Enabling Custom Generation of Novel Antibodies with Epitope Specificity and High Affinity, Precisely Meeting Specific Biological Functions, Expanding Applicable Antibody Targets, and Creating Value in a Broader Druggable Antibody Sequence Space.

Huang Xiaolu, Chairman and CEO of Clickmab, stated, "The results of this phase of the experiment validate the feasibility of AI generating antibody sequences from scratch, as well as the significant potential for model improvement with new antibody data. This represents a valuable exploration by XtalPi in simulating key aspects of the immune response using AI methods. Clickmab's digital antibody generation technology is not constrained by in vivo immune responses, offering a more efficient and innovative approach to targeting the 'high-hanging fruits' in antibody drug discovery."

Lai Lipeng, Co-founder and Chief Innovation Officer of XtalPi, stated, "Clickmab's breakthrough in AI antibody design effectively demonstrates the tremendous value that can be generated by combining highly relevant experimental data with AI. We are very pleased that XtalPi has been deeply involved in this endeavor. The XtalPi Innovation Center will also continue to explore breakthroughs and value transformation in areas including antibody design, vaccine development, preclinical, and clinical research, together with our partners."

Next, Clickmab will continue to invest in the development of the AI antibody platform, accumulate more high-quality data, further improve the efficiency and accuracy of predictive algorithms, expand its antibody programming capabilities, and achieve more precise de novo antibody design across a broader range.

About Clickmab

Clickmab is a biotechnology company focusing on the early discovery of antibody drugs. It was jointly funded by XtalPi and Biocytogen in June 2021. By utilizing self-generated high-quality human antigen-antibody pairing data, Clickmab has built the industry's first data-driven AI antibody de novo design platform targeting specific epitopes, which has been experimentally validated. This platform has achieved a breakthrough by generating antibody molecules that bind to specific epitopes from scratch using AI without relying on animal immunization experiments or antibody screening libraries. This innovation provides the industry with a third source for antibody discovery, significantly expanding the space for antibody drug discovery.

Clickmab is committed to building an AI and real-data-driven B-cell programming platform, exploring a new model for antibody drug development. By using AI algorithms to simulate and surpass the natural immune processes of the human body, it accelerates the design of highly targeted and development-friendly antibody molecules, filling the gaps in clinical medications. This effort aims to bring safer and more effective biologics to patients sooner.

About XtalPi

XtalPi is a technology company driven by intelligence and automation in drug research and development, providing integrated drug discovery solutions for global biopharmaceutical enterprises. XtalPi pioneered a new drug R&D model that combines intelligent computing, automated experiments, and expert experience, aiming to build a tripartite R&D platform. Through professional services, XtalPi helps customers shorten the drug R&D cycle and improve the success rate of drug development.

PharmaCube Pro

New Targets, New Technologies, New Therapies

Media Cooperation: 15895423126

Copyright © 2023 PHARMCUBE. All Rights Reserved.

Welcome to forward, share, and reasonably cite. When citing, please clearly indicate the source of the article in a prominent position; if you need to reprint, please leave a message to the WeChat Official Account backend or send a message, and include the name and ID of the official account.

Disclaimer: The information in this WeChat article is for general reference only and should not be directly used as decision-making content. PharmaCube assumes no responsibility for any loss incurred by any party due to the use of the content herein.