Numerate

Numerate, Inc. operated as a data-driven drug design company, specializing in the application of artificial intelligence and cloud-scale computation to transform small molecule drug discovery. The company was co-founded in 2007 by Guido Lanza, who served as CEO, and Brandon Allgood, who was the Chief Technology Officer. Their shared vision was to leverage AI to analyze vast amounts of drug discovery data, build predictive models, and ultimately accelerate the identification of new medicines.

Guido Lanza brought a background in molecular biology from UC Berkeley and bioinformatics from the University of Manchester, having previously co-founded another algorithm-focused biotech startup, Pharmix. This experience gave him a unique perspective on the scientific, technological, and business challenges of building an AI-platform-based company. Brandon Allgood, a Ph.D. in Theoretical Cosmology from UC Santa Cruz, applied his extensive experience in large-scale computational methods and machine learning to develop Numerate's core technology. His background in physics and software simulation was pivotal in creating the sophisticated algorithms that powered the platform.

The company's business model evolved from being a medicinal chemistry service provider to an AI-driven drug discovery engine that formed high-profile collaborations with large pharmaceutical companies like Merck and Lundbeck. These partnerships focused on identifying clinical candidates for a range of diseases, including cardiovascular and neurodegenerative disorders. Numerate's revenue was generated through a combination of upfront fees, R&D support, and milestone payments and royalties from these collaborations. The core of the business was its proprietary Data-Driven Drug Design (D4) platform. This platform combined machine learning, statistical modeling, and traditional medicinal chemistry to predict how a small molecule would behave in terms of activity, absorption, distribution, metabolism, and excretion (ADME) and toxicity. By running simulations on billions of potential compounds, the platform could rapidly identify promising candidates, aiming to reduce the multi-year timeline and high attrition rates common in preclinical drug discovery.

A significant milestone was the 2017 acquisition of DeepCrystal's small molecule design platform, which enhanced Numerate's capabilities with graph-valued deep-learning technology. In September 2019, Numerate itself was acquired by Valo Health, a life sciences company backed by Flagship Pioneering. The acquisition integrated Numerate's team and its platform—which by then had created over 30,000 models and powered more than 25 drug programs—into Valo's broader effort to redefine the drug development process.

Keywords: Numerate Inc, Guido Lanza, Brandon Allgood, AI drug discovery, computational drug design, small molecule therapeutics, machine learning in pharma, D4 platform, ADME modeling, predictive toxicology, computational chemistry, Valo Health acquisition, Merck collaboration, Lundbeck collaboration, preclinical development, drug design platform, artificial intelligence healthcare, computational biology, bioinformatics, deep learning drug discovery, pharmaceutical AI, CNS disorders, cardiovascular disease targets