Numerate Files IPO Prospectus Highlighting AI-Driven Drug Discovery Platform with DoD Collaboration

The lengthy development cycle, substantial investment, prolonged duration, and low efficiency of new drug development have long been a major concern in the pharmaceutical industry. According to data from the Tufts Center for the Study of Drug Development, it takes approximately 96.8 months for a new drug to go from research and development to approval by the U.S. FDA. Meanwhile, a research report by TechEmergence shows that the integration of artificial intelligence can increase the success rate of new drug development from 12% to 14%. This modest 2% increase can save the biopharmaceutical industry billions of dollars while significantly improving R&D efficiency.

NumerateFounded in 2007 and headquartered in San Bruno, California, USA, the company combines advanced developments in computer science and statistics with traditional chemical methods in drug discovery. It is dedicated to providing a drug design platform for enterprises developing small-molecule therapies. The platform primarily focuses on the research and development of drug candidates for cardiovascular diseases, metabolic disorders, neuropsychiatric conditions, Alzheimer’s disease, and Huntington’s disease.

NumerateDr. Guido Lanza, President and CEO, stated that Numerate has 11 years of R&D experience, setting it apart from other AI-driven drug discovery companies. Its algorithms can extract valuable insights from very small datasets and apply them to address novel biological challenges.

3D Ligand-Based Modeling Enables Machine Learning to Address Phenotype-Driven Drug Discovery Challenges Without Requiring Compound Structural Data. Such Discovery Efforts Often Involve Low-Throughput, High-Content Biological Problems.

2014In June, Numerate secured an $8 million Series C financing round, with Atlas Venture and Lily Ventures as the investors.

Guido LanzaThe PhD stated that this investment will advance Numerate’s further research in metabolic and cardiovascular diseases, aiming to improve conditions such as diabetes, obesity, heart failure, and arrhythmias. Numerate focuses on building data-driven drug design models to promote high-efficiency and predictable pharmaceutical analysis. This platform is particularly suitable for scenarios where data is scarce or where traditional methods are limited in generating the lead compounds required for drug development.

NumerateThe company’s team comprises employees from diverse fields, includingComputer scientists, data scientists, medicinal chemists, and biologists. The company brings together expertise from various fields to optimize the entire drug development process with the aim of building optimal solutions.

From left to right: Guido Lanza, Brandon Allgood, and John Griffin

According to VCBeat, Guido Lanza, President and CEO of Numerate, was named one of the “Top 30 Under-30 Tech Entrepreneurs” by BusinessWeek in 2006. Dr. Lanza previously co-founded Pharmix Pharmaceuticals, where he served as Chief Technology Officer and sat on the board of directors for six years.

Prior to joining Pharmix, Dr. Lanza, as a research scientist, collaborated with Professor Koza at Stanford University to develop genetic programming applications in bioinformatics and computational biology. Furthermore, Dr. Lanza is the author of ten scientific publications and the inventor of four granted patents. He holds dual bachelor’s degrees in Molecular and Cell Biology and Integrative Biology from the University of California, Berkeley, as well as a Master’s degree in Bioinformatics from the University of Manchester in the United Kingdom.

NumerateChief Technology Officer Brandon Allgood is primarily responsible for managing the company’s AI platform and applications, as well as charting the blueprint for its future technological development. He holds a Ph.D. in Computational Physics from the University of California, Santa Cruz, and has previously served as the company’s Director of Computer Science and as a Research Scientist at Pharmix.

AllgoodHe has authored scientific books on astrophysics, solid-state physics, and computational biology, and possesses 15 years of professional experience in large-scale cloud computing, distributed computing, artificial intelligence, and mathematical modeling. Additionally, he is a member of the Forbes Technology Council and a trustee of the UCSC Foundation.

As Chief Strategy Officer of Numerate, Inc., John Griffin oversees the company’s therapeutic programs and project collaborations. He was also a co-founder and former Chief Scientific Officer of Theravance, Inc., a publicly traded biopharmaceutical company. During his tenure as an Assistant Professor at Stanford University, Griffin published 39 scientific papers.

As the inventor of 27 patents, he has received numerous awards, including the American Chemical Society’s Arthur C. Cope Scholar Award and Stanford University’s Dean’s Award for Teaching. In addition, he has served as a member of the Wellcome Trust’s Fundraising Committee, was a National Science Foundation Postdoctoral Fellow at Harvard Medical School, and earned his Ph.D. in Chemistry from the California Institute of Technology.

NumerateLanza, CEO and President, stated that the company’s proprietary ADME and toxicity prediction platform is a key pillar of its business.

ADMEADME is the abbreviation for "absorption, distribution, metabolism, and excretion," describing the disposition process of a drug compound within a living organism. These four parameters determine drug concentration and the kinetics of tissue exposure; therefore, ADME can be used to evaluate the compound's performance as a drug and its pharmacological activity.

Traditional biotechnology and pharmaceutical startups typically focus on a limited number of targets or a single therapeutic area, whereas new drug design platforms primarily leverage machine learning techniques to simulate the pharmacological properties of small-molecule compounds, such as target binding affinity and specificity, pharmacokinetic and metabolic profiles, and toxic side effects. The platform’s standard drug screening workflow employs specific models for drug activity, specificity, and ADME (absorption, distribution, metabolism, and excretion) to select 25 million compounds from a library of one trillion virtual compounds for in silico testing.

Studies have shown that the entire process can be completed in just one week, with each simulated compound tested at a cost of $0.0001. Chemists analyze the test results to select the most promising simulated compounds for synthesis and experimental validation.Experimental results are used to refine and improve the accuracy of simulations; as this process iterates, the candidate compounds generated by the simulation system become increasingly targeted.Thus, it is evident that Numerate’s pharmaceutical platform can leverage machine learning to address biological challenges with high efficiency and cost-effectiveness, utilizing the platform to focus on and validate data, thereby improving the entire pharmaceutical industry.

Additionally, Lanza stated that they have secured over $10 million in investment in this field, including a major contract with the U.S. Defense Threat Reduction Agency (DTRA) to establish and validate a system capable of rapidly transforming lead compounds into clinical candidates. Many of Numerate’s collaborations with pharmaceutical companies are based on its ADME and toxicity prediction capabilities. What sets it apart is its ability to learn from all past R&D projects, thereby providing accurate decision-making support for every future chemical design and candidate selection.

2017In June, Takeda Pharmaceutical Company officially signed an agreement with Numerate, Inc. to collaborate on the development of small-molecule drugs for oncology, gastroenterology, and central nervous system disorders using Numerate’s AI technology. Lanza stated that Numerate aims to establish partnerships with large pharmaceutical companies in the form of product line acquisitions, and Takeda will receive upfront licensing rights for candidates generated by Numerate’s AI platform.

NumerateDr. Lanza, the company’s CEO, predicts that AI technology will be applied across the entire pharmaceutical industry within the next three to five years. In his view, data technology and business value are crucial to operating an AI startup.

Understanding algorithms and their applications in the biological field, as well as the data to which these algorithms are applied, is key to creating experimental models. Lanza stated that to address these issues, they spent decades understanding the data, overcoming technical barriers, navigating the complexities of biology (and its associated noise), and tackling chemical challenges (along with the biases they introduce).

From a business perspective, Numerate places greater emphasis on customer needs. Dr. Lanza stated that their new drug R&D team focuses on the practical value of their products in both commercial and academic sectors. Their revenue model stems from manufacturing and scaling up disease-treatment-related products, with the ultimate goal of licensing the company’s computational programs and applied technologies. Their applications fall into the following two categories:

Collaboration Program Schedule

NumerateA portion of the investment strategy involves collaborations with other companies, including the joint development of clearly defined, aligned product plans with partners, such as lead compound optimization projects. These optional programs pertain to the target classes and therapeutic areas developed by Numerate. The key to these collaborations lies in generating revenue through a co-created, externalized research model.

Internal Procedures

NumerateThe second part of the investment includes internal projects developed with a focus on licensed technologies. This also covers technology products and designs purchased through collaborations with other companies. For example, in June 2017, Numerate co-acquired small-molecule modulators of ryanodine receptor 2 (RyR2) with Servier, an international pharmaceutical company headquartered in France. These modulators target cardiovascular-related diseases, aiming to improve symptoms such as heart failure or arrhythmias in patients. Servier primarily drives corporate development through innovation in five key areas: diabetes, cancer, immune-inflammatory diseases, neurodegenerative diseases, and its promotion and activities in the high-quality generic drug sector.

1. Comparison of Numerate and Recursion Pharmaceuticals 

Recursion, founded in 2013, has an ambitious goal: to discover treatments for 100 diseases within a decade. The company’s core technology leverages computer vision to process cellular images and evaluates the effects of drug administration on diseased cells by analyzing more than 1,000 cellular features. By employing advanced imaging and artificial intelligence technologies, this platform enables high-throughput experiments using cellular models, allowing for the screening of thousands of candidate drugs across cellular models of hundreds of diseases. Recursion has already identified 15 candidate drugs for rare diseases, with one candidate for the treatment of cerebral cavernous malformation poised to enter clinical trials.

In contrast to Numerate, Recursion Pharmaceuticals leverages AI technology to conduct in-depth analysis of cellular structures, primarily promoting novel drug discovery through image-based analysis. Its technological principles and research directions differ from those of Numerate; Recursion focuses on rare diseases and genetic disorders, whereas Numerate is dedicated to explorations in oncology, gastroenterology, and central nervous system diseases.

2.  Comparison between Numerate and Insilico Medicine

Insilico Medicine, founded in 2014, is dedicated to extending human healthspan. To this end, the company has collected extensive multi-omics data from healthy and diseased individuals across various age groups. It employs machine learning to comprehensively analyze these data, identifying biomarkers associated with aging and disease, repurposing existing marketed drugs, and discovering novel anti-aging therapeutics. Another core business line involves collaborating with research institutes and pharmaceutical companies, leveraging its expertise in deep neural networks and machine learning to assist in drug discovery, biomarker identification, and the development of new tools for aging research.

A Comparison of the Two Companies: Numerate and Insilico Medicine Share Similar Approaches in Drug Molecule Analysis. Both employ general machine learning, data analytics, and experimental simulation to conduct in-depth investigations of molecules. Compared to Insilico Medicine, Numerate has a longer history and has raised more funding. However, Insilico Medicine has performed no less impressively in terms of collaborative projects and product output. According to VCBeat, the company has established partnerships with over 150 institutions worldwide.