GIP (glucose-dependent insulinotropic polypeptide) and GLP-1 (glucagon-like peptide-1) both belong to the incretin hormone family, and their receptor, GIPR, plays a multidimensional, central role in human metabolic regulation: it promotes lipid uptake and storage in adipose tissue; it regulates bone formation and turnover in the skeleton; and it participates in appetite regulation within the central nervous system (CNS). Recently, INSILICO announced the discovery of an innovative GIPR antagonist, ISM0676, which demonstrated a powerful efficacy of 31% weight loss over 27 days in a preclinical mouse model, providing new preclinical data to evaluate the role of the GIPR pathway in weight loss therapy. It was disclosed that ISM0676 was designed with the assistance of INSILICO's (03696) Pharma.AI artificial intelligence drug discovery platform, taking just 14 months from project initiation to the nomination of a preclinical candidate compound, with a cumulative synthesis and testing of no more than 200 molecules, showcasing its exploratory path of using AI to enhance early-stage R&D efficiency. So, will GIPR antagonists become the next "miracle drug" pursued by the industry? And how significant a role will artificial intelligence play in areas such as mechanism exploration and molecular optimization?

In the rapidly advancing weight loss drug market, GLP-1, a target discovered decades ago with a relatively mature mechanism, has taken the lead into clinical application, attracting startups, pharmaceutical giants, and even cross-industry players to enter the field. In 2025 alone, the sales of GLP-1 products from global top pharmaceutical companies like Eli Lilly and Novo Nordisk have already exceeded $50 billion, highlighting the immense clinical demand and commercial potential of the weight loss赛道. However, as the user population expands and follow-up periods lengthen, certain limitations of GLP-1 therapy in practical application have drawn increasing attention, such as the reduction in muscle mass, where approximately 40% of the weight loss composition comes from the loss of lean body mass, leading to a decrease in basal metabolic rate; the efficacy plateau, where patients often face a weight loss stall in the second year of treatment, making further weight loss difficult even with continued medication; weight rebound after discontinuation, as the underlying fat storage signals are not fundamentally altered, often leading to rapid weight regain and adipose tissue rebound; and the impact of long-term injection and gastrointestinal adverse effects on patient compliance. Against this backdrop, the market has begun seeking solutions that not only focus on weight loss intensity but also improve safety and compliance. Some R&D directions are turning to combination strategies involving GIPR antagonists and GLP-1R agonists, hoping to achieve weight loss while effectively preserving muscle mass.

Against the backdrop of the aforementioned demand, INSILICO chose GIPR as an entry point to advance the early discovery and optimization of related candidate molecules. Supported by its Pharma.AI platform, it utilized AI application tools such as the Chemistry42 generative chemistry engine and the Alchemistry free energy binding prediction engine to conduct molecular design and binding mode prediction. In early-stage R&D, it simulated the detailed interactions between drug molecules and the GIPR receptor, thereby assisting in the optimization of metrics like novelty and druggability, and reducing the risk of drug-drug interactions (DDI). The company stated that this process, which prioritizes computational screening followed by targeted experimental validation, helps enhance iterative efficiency and R&D precision, while keeping early-stage research within a relatively limited scale of molecular synthesis and testing. Preclinical research data showed that in a diet-induced obesity (DIO) humanized GIP mouse model, the combination of ISM0676 with semaglutide achieved approximately 31.3% weight loss over 27 days; the control semaglutide monotherapy group achieved approximately 16.2%. Furthermore, the ISM0676 monotherapy group also showed approximately 10.4% weight reduction in this model. These results provide further preclinical evidence for the "GIPR antagonist plus GLP-1" combination strategy. It was disclosed that ISM0676 demonstrated relatively positive preclinical characteristics in areas such as in vivo metabolic stability, risk assessment of potential drug interactions, safety, and predicted clinical dose, providing support for its subsequent development. The company also believes that this molecule holds potential value in improving body composition (such as the proportion of lean body mass) and that its combination with GLP-1 regimens may yield synergistic effects; however, these conclusions still require validation in further research and clinical trials.

Previously, Dr. Alex Zhavoronkov, founder and CEO of INSILICO, stated that the field of metabolic diseases (Cardiometabolics) holds the potential to become the first global breakthrough area for achieving mass healthspan extension. There is also academic theoretical basis for this: for instance, a 2023 paper published in the academic journal Cell listed nutrient sensing dysregulation, mitochondrial dysfunction, and long-term chronic inflammation as part of the "hallmarks of aging" framework, providing a researchable perspective on the underlying biological mechanisms of common health issues like obesity, type 2 diabetes, and cardiovascular diseases. According to INSILICO, its AI-driven R&D process covers stages from large-scale data and pathway analysis and target discovery to generative molecular design and multi-property prediction and optimization, and it has already established a pipeline of a certain scale: there are currently about 27 projects in clinical or preclinical stages, including what it describes as its rapidly progressing AI-driven FIC project Rentosertib, with several partnered or self-developed projects having also entered Phase I clinical trials. As projects like ISM0676 demonstrate certain potential for weight loss and body composition improvement in preclinical data, the application boundaries and practical output of artificial intelligence in metabolic disease drug R&D are also expected to become clearer through subsequent research and clinical validation.