Lilly Partners with Atomwise to Accelerate Drug Discovery Using AI-Powered Screening Technology

VCBeat (WeChat Official Account: vcbeat) has learned that medical technology company Atomwise recently announced a multi-year collaboration agreement with Eli Lilly. Under the agreement, Eli Lilly has obtained a license to use Atomwise’s proprietary AI-driven drug discovery technology for preclinical drug development.

The two companies may also collaborate on up to 10 drug targets selected by Eli Lilly. Upon completion of milestone events for each target, Atomwise will receive $1 million in funding and will be eligible for potential development and commercialization milestone payments totaling up to $550 million. As part of the agreement, Atomwise may develop drugs for collaborative targets that Eli Lilly elects not to advance into clinical testing.

Atomwise, founded in June 2012 in San Francisco, USA, is a company that leverages supercomputing for drug research and development. The company aims to simulate pharmaceutical processes using supercomputers, artificial intelligence (AI), and complex algorithms to predict the efficacy of new drugs and reduce R&D costs. Atomwise’s primary research focus is on utilizing AI to accelerate compound screening, thereby facilitating new drug discovery.

Atomwise pioneered AtomNet, the first structure-based small-molecule drug discovery platform. Convolutional neural networks achieve state-of-the-art predictive performance in fields such as speech and image recognition by hierarchically combining simple local features into complex models. AtomNet was developed by applying this approach to chemistry to predict the bioactivity of small molecules against therapeutic targets.

The AtomNet platform, based on convolutional neural networks, extracts insights from hundreds of experimental affinity measurements and thousands of protein structures to predict the binding of small molecules to proteins. This platform enables chemists to identify targets with high precision and accuracy, conduct lead optimization, and perform toxicity predictions.

AtomNet can autonomously learn the features governing molecular binding, thereby avoiding the typical manual tuning and parameterization of binding features required in traditional computational methods. By applying local convolutional filters to structural information of protein targets, AtomNet is capable of successfully predicting novel active molecules for these targets.

Today, Atomwise has partnered with leading pharmaceutical and agrochemical companies, as well as over 100 universities and hospitals across 19 countries, to carry out hundreds of projects.

Eli Lilly and Company is a global, research-based pharmaceutical company headquartered in Indianapolis, Indiana, USA, dedicated to providing innovative, medicine-based healthcare solutions for people worldwide.

Dr. Abraham Heifets, CEO of Atomwise, stated, “Eli Lilly has clearly indicated that they will focus on developing drugs targeting novel protein targets, as these areas are often under-researched and present significant challenges. Eli Lilly aims to leverage Atomwise’s AI-driven drug discovery technology on its virtual molecular library and integrate it with the automated synthesis capabilities of its robotic laboratories. Our expertise and tools have proven successful in achieving these objectives, positioning our technology as a key enabler for improving patient health.”

(Compiled by: Li Chengping)