Insilico Medicine Files IPO Prospectus Highlighting World’s First AI-Discovered Novel Mechanism Drug for Idiopathic Pulmonary Fibrosis
Insilico Medicine has secured multiple industry firsts in the biotechnology and drug discovery sectors for achieving the world’s first AI-driven discovery of a novel-mechanism drug for idiopathic pulmonary fibrosis:
• Through successful validation in multiple human cell and animal model experiments, the efficacy and safety of new drug targets and molecules developed by artificial intelligence have been demonstrated. Today’s breakthrough marks the first scientific validation of AI in the industry, applying it to new drug development through to the preclinical study of candidate compounds.
• First integration of biology and chemistry via artificial intelligence in the new drug development process: Historically, the steps of identifying new targets, designing novel compounds, and validating their efficacy through preclinical and clinical studies have been distinct phases in the AI-driven drug discovery process.
• Insilico Medicine has broken records for speed and lowest cost in preclinical candidate selection—significantly accelerating and advancing preclinical development while saving millions of dollars in drug development costs.
Insilico Medicine Achieves Breakthrough in AI and New Drug Development—First to Combine Biology and Generative Chemistry to Discover a Novel Clinical Candidate for Idiopathic Pulmonary Fibrosis (IPF) with a Unique Mechanism of Action, Successfully Validated Through Multiple Human Cell and Animal Model Experiments. IPF involves multiple pathologies affecting various organs (lungs, liver, and kidneys), and the emergence of this new drug holds promise for addressing the broad unmet medical needs impacting hundreds of thousands of people worldwide.
The etiology of idiopathic pulmonary fibrosis (IPF) remains unclear, and its pathogenesis is not yet fully understood by the medical community. The disease is predominantly sporadic, with a median survival of no more than five years from symptom onset to death.
Extensive pulmonary fibrosis is prone to complicating into lung cancer, and pulmonary hypertension may also develop in the late stages. Drugs currently used to treat IPF have been in clinical use for over 30 years, but they are effective in only 10%–30% of patients. In the late stages of the disease, patients rely on oxygen therapy to improve their quality of life, but the prognosis remains poor.
Dr. Alex Zhavoronkov, Founder and CEO of Insilico Medicinestating, “Linking the right drug targets to the right diseases is the greatest challenge in drug development.” “As we reach the milestone today of achieving the first AI-discovered and scientifically validated preclinical candidate (PCC), Insilico Medicine has overcome another major obstacle in drug discovery and broken through another bottleneck in the traditional drug discovery process, accomplishing this at significantly reduced cost and time.”
From target discovery to the invention of preclinical candidate drugs, Insilico took less than 18 months to achieve target identification, molecular generation, and validation through traditional experiments, including confirmation of efficacy in animal models of idiopathic pulmonary fibrosis (IPF) and safety assessment. The total cost was approximately $1.8 million, with an additional $800,000 spent on efficacy studies for other fibrotic diseases. No more than 80 small-molecule compounds were synthesized and tested.
Traditional drug discovery begins with the screening of tens of thousands of small molecules, followed by the synthesis and testing of hundreds more to identify a handful of candidates suitable for preclinical studies. Of these, only about one in ten ultimately passes clinical trials in human patients. The entire process is slow and costly, averaging 10 years and requiring billions of dollars in expenditure.
Another barrier that further hinders the launch of new drugs to the market is that the extensive R&D steps involved in the entire research and development process—each stage costing hundreds to tens of millions of dollars—are often carried out in a fragmented manner by different companies or distinct business units within the drug development industry.
Dr. ZhavoronkovIndicates:
“We are rewriting the history of drug discovery, becoming the pioneer and leader of the first—and only—AI-driven integrated drug discovery system.” “By creating the first universal system that connects all areas of drug development, from target identification and small-molecule compound design to future clinical trial outcome prediction, Insilico’s AI platform will be able to support advancement at every stage of drug R&D.”
Insilico Medicine began its research with 20 novel potential targets associated with fibrosis discovered through artificial intelligence, gradually narrowing the therapeutic focus to a single new target specifically for idiopathic pulmonary fibrosis (IPF).
Following target identification, Insilico Medicine employed its AI-driven chemistry generation system to design a novel set of compounds aimed at selectively inhibiting this new target. These molecules were required to exhibit favorable selectivity, bioavailability, metabolic stability, oral druggability, safety, and multiple other high-quality attributes characteristic of promising drug candidates. Initially generated by the structure-based molecular design algorithms within Chemistry42, the company’s generative chemistry AI platform, these molecules demonstrated efficacy in both cellular assays and animal models.
These experimental data are then fed back into the artificial intelligence system, which redesigns a new batch of compounds to optimize their activity and druggability, followed by further validation.
After several rounds of design–synthesis–evaluation–optimization–redesign cycles, a preclinical candidate compound has been identified. Insilico’s preclinical candidate has undergone rigorous evaluation by both internal experts and external specialists in the field of fibrotic diseases, and has now entered the preclinical research phase.
Furthermore, the company has used artificial intelligence to predict a high probability of success for Phase II clinical trials of this new IPF target and novel molecule. Insilico is currently conducting IND-enabling studies, with the goal of initiating clinical research in early 2022.
Insilico welcomes and looks forward to collaborating with pharmaceutical companies on post-Phase II drug development.
Although hot topics surrounding new drug development typically focus on when new targets are discovered or when new drugs enter clinical trials, the area currently most conducive to innovation and with the greatest business impact lies between target discovery and clinical development.
In 2019, Insilico Medicine made history by inventing and launching a novel artificial intelligence system for drug discovery, capable of generating entirely new molecules from start to finish in just 21 days at a cost of approximately $150,000. Given that the failure rate in target discovery stands at around 95%, Insilico addressed one of the most significant bottlenecks in the pharmaceutical industry’s drug discovery process at the time. Powered by generative chemistry leveraging modern AI technologies, Insilico’s AI software can rapidly generate novel molecular structures with specific desired properties.
As the first company to explore the use of generative adversarial networks (GANs) and generative reinforcement learning (RL) artificial intelligence technologies for drug discovery, Insilico’s AI software has provided scientific validation to the industry by demonstrating the first successful identification and generation of novel clinical candidate compounds.
Dr. Zhavoronkov stated:
“The pinnacle of the deep learning revolution can be traced back to 2014, when Generative Adversarial Networks (GANs) emerged and deep learning systems began to surpass human performance in image recognition. The company was founded in the same year. In 2016, we experimentally validated that deep learning systems could identify novel biological targets from omics data. From 2017 to 2019, we continuously demonstrated that generative artificial intelligence could invent and design new molecules with activity in human cells and animal models.”
However, a major challenge remains: Can artificial intelligence design a novel molecule for a new target that has no known inhibitors and has not been validated in disease contexts? We have now successfully integrated biology and chemistry to nominate a preclinical candidate drug capable of acting on a new target, with the aim of advancing it into human clinical trials. This represents a far more complex and higher-risk challenge, greater by an order of magnitude, that urgently needs to be addressed.
“As far as I know, this is the first case in which artificial intelligence has successfully identified a novel target and designed a preclinical drug candidate capable of addressing disease indications affecting large populations. This represents a significant milestone for us. Our ultimate ‘moonshot’ is to tackle human aging, which requires more advanced and reliable AI technologies to help us understand and modulate human biology in other chronic diseases.”
Furthermore, Insilico Medicine has secured substantial funding to support drug discovery and development across multiple novel therapeutic targets. The company already provides target discovery and generative chemistry systems and services to pharmaceutical and biotechnology firms through its proprietary Pharma.AI software suite. Its AI-driven target discovery platform, PandaOmics, is available as a software-as-a-service (SaaS) offering, while its small-molecule generative chemistry platform, Chemistry42, began on-site installation and deployment at pharmaceutical enterprise customers in September 2020. To date, the world’s most advanced pharmaceutical companies have started adopting our Chemistry42 molecular generation and design platform, and PandaOmics has been deployed within the drug target discovery divisions of numerous prestigious academic institutions and pharmaceutical companies.
Insilico also announced that it will continue to expand its scientific research team. The company has established a team in Shanghai comprising more than 20 senior drug development professionals, led by Dr. Ren Feng, Chief Scientific Officer (CSO), who joined Insilico in February this year. Previously, Dr. Ren served successively as Senior Vice President of the Biology and Chemistry Departments at Medicilon Biopharma and as Chemistry Director at GlaxoSmithKline (GSK). This team is responsible for advancing new drug candidates discovered through artificial intelligence into clinical trials and building a broad preclinical/clinical drug portfolio.
The software developed by Insilico Medicine leverages generative models (GANs), reinforcement learning (RL), and other modern machine learning techniques to generate novel molecular structures with specific properties. Insilico Medicine has also developed software for molecule generation, target identification, and prediction of clinical trial outcomes. The company integrates two business models: providing AI-driven drug discovery services and software through its self-developed Pharma.AI platform (www.insilico.com/platform/), and independently developing preclinical and clinical programs. Preclinical projects are realized by identifying new drug targets and novel molecules via its proprietary platform. Since its establishment in 2014, Insilico Medicine has raised over $52 million in funding and received multiple industry awards. The company has also published more than 100 peer-reviewed papers and filed over 25 patents.