Amid the rapid advancement of AI, its application in drug development is gradually transforming the traditional pharmaceutical industry, driving drug research and development towards faster, stronger, and more precise innovation. As a leader in the AI pharmaceutical field, Insilico Medicine has demonstrated impressive performance over the past few years. Whether in pipeline development, platform technology, or international collaborations, it has showcased its exceptional capabilities.
Can AI-driven drug development ultimately lead to success in new drug research? Can it effectively accelerate the drug development process through algorithms and data? As a partner of Insilico Medicine, Medicilon is honored to invite Dr. Feng Ren, the joint CEO and Chief Scientific Officer of Insilico Medicine, to discuss and interpret the developments and trends of AI in new drug research and development.
Dr. Feng Ren, joint CEO and Chief Scientific Officer of Insilico Medicine, graduated from the Department of Chemistry at Harvard University and has over 15 years of experience in the drug development industry. He previously served as the Head of Small Molecule Innovative Drug Development at the multinational pharmaceutical company GSK. In 2018, he joined Medicilon, where he has held positions as Vice President and Senior Vice President. He has been fully responsible for the R&D service operations of the Chemistry and Biology departments, as well as the company's drug discovery platform services.
In 2021, Dr. Ren joined Insilico Medicine as Chief Scientific Officer, leading the drug development team in using the company's independently developed artificial intelligence platform to build preclinical and clinical product pipelines and manage external project collaborations. In June 2022, Dr. Ren was appointed as the joint CEO of Insilico Medicine. Dr. Ren has published over 30 international academic papers and holds more than 20 international patents.
Dr. Feng Ren: The occurrence rate of MTAP deletion in tumors is as high as 15%, making it one of the most common genetic deletions in cancers such as lung cancer, bladder cancer, and pancreatic cancer. It is associated with poor prognosis in patients. By targeting MAT2A, we hope to regulate key molecules that influence cell function and survival, thereby exerting selective anti-proliferative effects on MTAP-deficient cancer cells.
ISM3412 is an orally available small molecule inhibitor designed with assistance from the Chemistry42 synthetic chemistry platform, demonstrating potential as a "best-in-class" agent within its category. From preclinical data, ISM3412 shows excellent pharmacokinetic properties, potent efficacy at low doses, good solubility and permeability, as well as favorable safety characteristics.
Furthermore, ISM3412 not only holds potential as a monotherapy but also demonstrated synergistic effects when used in combination with standard therapies like chemotherapy in preclinical studies. Compared with known drug candidates with the same target, ISM3412 has a superior therapeutic window. Existing data supports the potential of this small molecule in diseases such as non-small cell lung cancer, mesothelioma, esophageal cancer, and bladder cancer.
Dr. Feng Ren: Since 2021, with the support of artificial intelligence platforms, we have internally developed over 30 pipelines and nominated 18 preclinical candidates (PCCs). Currently, 7 compounds have received clinical trial approvals. These projects span areas related to fibrosis, oncology, autoimmune diseases, and others.
From the perspective of our collaborative projects, two recent collaborations with Fosun Pharma are progressing steadily. Among them, the QPCTL project in the clinical stage, which focuses on oncology, has recently administered its first patient dose. Another synthetic lethality project has also received recognition from Fosun Pharma, with both parties agreeing on the nomination of the preclinical candidate compounds (PCCs), marking successful delivery of outcomes in our collaboration.
Looking at our in-house projects, this year we have nominated a preclinical candidate compound (PCC) targeting KIF18A. It is a potential best-in-class small molecule inhibitor targeting chromosome instability, with the potential to treat advanced solid tumors with TP53 mutations. On the clinical side, the MAT2A project has completed IND-enabling studies and obtained clinical trial approvals in both China and the United States. Our core pipeline IPF project recently completed enrollment of all patients in its Phase 2a clinical trial in China. We anticipate disclosing interim results and data later this year. These are all very exciting preclinical and clinical milestones for us.
Dr. Feng Ren: The areas we currently focus on have two characteristics. One is ample data availability, which fully showcases the capability of artificial intelligence to uncover clues and screen potential possibilities, an area where AI excels. Another characteristic is that the target molecules of the projects we are currently initiating are relatively novel. These projects not only possess significant market potential, avoiding homogenized competition, but also have the potential to break through in areas with unmet medical needs, such as rapidly progressing malignant tumors.
Dr. Feng Ren: After AI applications like ChatGPT exploded in popularity among consumer-facing (C-end) uses, generative AI and large language models (LLMs) seemed to enter various industries overnight. However, these technologies have actually been applied in pharmaceutical research and development for much longer. Insilico Medicine published the first paper in the field utilizing Generative Adversarial Networks (GANs) to empower innovative small molecule generation. Building upon this foundation, they extended upstream and downstream, establishing and validating the Pharma.AI platform across three major stages of drug development, spanning biopharmaceutical and clinical development domains.
In the traditional drug development field relying on researchers' expertise, challenges such as target redundancy and high costs with low efficiency persist. Having worked in frontline research and development for over a decade, I deeply resonate with this, and AI may indeed be the breakthrough tool. Unlike humans, AI can learn rapidly and continuously iterate and improve without limits. Insilico Medicine's proprietary database alone covers over 1 billion data points, surpassing the lifetime literature that a single scientist could ever read. Combining generative AI, natural language processing, Transformers, and other cutting-edge AI technologies, we expect AI to play a comprehensive role in reducing costs and increasing efficiency across the pharmaceutical research and development field.
Dr. Feng Ren: The Pharma.AI platform and algorithms are constantly undergoing continuous expansion and optimization, typically with minor updates every few months and major updates every six months. Currently, the platform covers over 30 algorithms, all selected and refined through rigorous testing and optimization. In November last year, we released an update for Pharma.AI, expanding the database, refining data labeling logic, enhancing real-time question answering using natural language processing, and adding functionalities like ADMET property prediction. In the future, we expect AI technology to empower not only biopharmaceutical but also various other fields. We are currently exploring green chemistry and new materials, and have achieved international collaboration in these areas.
Dr. Feng Ren: In the AI pharmaceutical track, China started relatively late, with most companies being founded around 2016-2018, lagging behind by several years. For this reason, it's not surprising that overseas AI pharmaceutical companies have been the first to advance drug pipelines into Phase 2 clinical trials. However, leveraging high-quality domestic talent and the accumulated advantages in biopharmaceutical resources, we may have the potential, and perhaps have already achieved, overtaking on the curve. For example, Insilico Medicine's leading AI-discovered target for anti-fibrotic small molecules, with molecular structures designed by AI, administered to the first group of patients in Phase 2 clinical trials last June, has become the world's first true "AI drug" to achieve this milestone.
Dr. Feng Ren: I believe in the potential of AI pharmaceuticals. Traditional drug development relies on human knowledge and experience, which can reach a plateau that is difficult to surpass. But AI is different; it has no human biases and can sometimes surprise us with brilliant insights. Moreover, AI will continue to progress with increasing amounts of data, potentially without limits.
Currently, AI primarily empowers drug development in the early stages, known as preclinical phases. This includes target discovery, molecular design, synthesis route simulation, crystal form and salt form prediction, ADMET property prediction, and more. As data standardization and openness in development progress, we can expect AI to play a role in areas it currently cannot empower, such as predicting toxicity properties. In the clinical stage, we can also expect AI to demonstrate its value in screening eligible patients for enrollment and potentially even in managing control groups for clinical trials.
Dr. Feng Ren: Although globally AI has enabled the research of over 100 drugs, overall, AI pharmaceuticals are still in a relatively early stage of development. Most are concentrated in Phase 1 clinical trials, with the fastest progressing to Phase 2; there have been no AI-discovered drugs approved for market yet. AI pharmaceuticals have not yet received conceptual validation, so industry skepticism and cautious investment are quite normal.
We believe that developing a suitable business model that fits its own needs before AI completes full validation of the entire process is the primary challenge that the entire AI pharmaceutical sector needs to confront. Insilico Medicine is also continuously exploring, from software partnerships and pharmaceutical research and development services to licensing out its drug pipelines. But no matter how good the business model is, if it cannot bring benefits to the company, it will eventually be unsustainable.
In 2023, Insilico Medicine successfully completed external licensing agreements with Exelixis and Menarini, receiving substantial upfront and milestone payments, leading to a significant increase in revenue for the year. Based on this experience, we believe that external licensing of pipelines could be a crucial business model for the company to achieve self-sustainability.
Dr. Feng Ren: My experience in the CRO industry was very helpful. During my over ten years at GSK, I primarily worked in drug chemistry research and development. After joining Medicilon, I first led the research and development services in the chemistry department. Later, I took over the R&D services in the biology department and Medicilon's drug discovery platform services. It provided me with the opportunity to be involved in various stages of early drug discovery, including biology, chemistry, CMC, toxicology research, and more. This gave me a comprehensive and hands-on understanding of the end-to-end drug discovery process, which has laid a strong foundation for me to contribute effectively to clinical-stage drug discovery at Insilico Medicine.
In addition to improving my personal capabilities, my experience at Medicilon enabled me to fully understand and trust the capabilities of CROs. After joining Insilico Medicine, I have become more accustomed to and engaged in extensive collaborations with CROs. Currently, Insilico Medicine has collaborated with over 40 CROs globally, including Medicilon. Our approach involves using an artificial intelligence platform for virtual screening. Subsequently, experienced biologists and medicinal chemists provide 2-3 rounds of expert feedback, followed by wet lab experiments conducted by CROs for validation. Our extensive CRO collaboration network allows us to advance projects rapidly, promptly, and efficiently in a capital-light manner.
About Insilico Medicine
Insilico Medicine is a clinical-stage biotechnology company driven by generative artificial intelligence. Through next-generation AI systems connecting biology, chemistry, and clinical trial analysis, Insilico Medicine utilizes modern machine learning technologies such as deep generative models, reinforcement learning, and transformer models to build a powerful and efficient AI drug development platform. This platform identifies novel targets and generates candidate drugs with specific molecular structures. Insilico Medicine focuses on areas of unmet medical needs such as cancer, fibrosis, immunology, central nervous system diseases, and age-related diseases, advancing and accelerating the development of innovative drugs.
Relevant
news
marketing@medicilon.com
