Relation Therapeutics, Active-Graph Machine Learning Drug Discovery Company, Appoints David Roblin as CEO and Lindsay Edwards as CTO
LONDON–(BUSINESS WIRE)–Relation Therapeutics, the pioneer in drug discovery using active graph machine learning (ML) combined with single cell analysis and deep clinical insights, announced today that it has raised a $25 million investment and hired a Managing Director and a Chief Technology Officer. The funding and expanded leadership will allow Relation to expand its platform and advance therapeutic candidates for diseases with few or no good drugs.
Dr. David Roblin joins the company as Chief Executive Officer and Dr. Lindsay Edwards as Chief Technology Officer and President of Platform. “We will dramatically improve the success of drug discovery – we have to, because patients are waiting for breakthrough drugs,” Roblin said. “Our approach brings together, for the first time, human genetics, single-cell omics, functional genomics and active graph machine learning in a single engineering design. We believe our techniques will dramatically accelerate the discovery of new, life-changing drugs. »
The critical bottleneck in drug discovery remains the poor understanding of the underlying biology of disease. As a result, we don’t know why patients get sick; and far too often, most drug candidates fail in trials and many devastating diseases remain untreated. Historically, successful drugs have mostly been discovered by sheer luck. Relation offers a radically different approach which makes it possible to better understand the biology of the disease and to rationally discover new therapies.
Relation’s platform uses the power of active graph ML, called Metagraph. The technology has been used successfully by tech companies to solve computer vision and product recommendation problems, but never before in drug discovery on this scale. With active graph machine learning, Relation can understand the large number of combinatorial functional relationships between genes, proteins and drugs. Relationship will drive some of this work with NVIDIA, as one of four startups to have access to the NVIDIA Cambridge-1 GPU supercomputer.
Relation’s pioneering “lab in the loop” embeds active learning into every step of drug discovery, from predicting cellular states to validating new targets. A significant challenge for any therapeutic enterprise using machine learning is that of “ground truth data,” or information known to be true. Working from real cells provided by proprietary biobanks, Relation’s technology generates genomic data that provides direct insights into critical biological relationships that feed directly into its ML systems. The platform then automatically requests new experiments to improve its predictive ability, cutting through otherwise intractable combinatorial space.
Relation initially focuses on bone disease, where there is a compelling unmet medical need, and because the company’s platform has a clear advantage where there is good cellular representation. “Relation brings a powerful and proven approach to interrogating extremely complex information to effectively formulate high-quality recommendations. The speed at which they operate and the progress they have made are remarkable and can be transformative. We are thrilled to join them in their efforts,” said Jason Pontin, DCVC Partners and Relations Council Member.
Christine Aylward, Founder and Managing Partner of Magnetic Ventures and Relation Board Member, said, “Beyond its breakthrough platform technology, the Relation team has demonstrated its ability to form partnerships and to attract an interdisciplinary set of world leaders in the field of machine learning, single-cell technology, genomics and drug discovery. We are delighted to partner with Relation and the investor syndicate to support the company’s vision to transform drug discovery through its innovative approach.
Dr. David Roblin is a leading scientist, physician, entrepreneur and leader in the life sciences industry. Roblin has been involved with Relation since its founding, while also being engaged as CEO of Juvenescence Therapeutics, one of the earliest investors in Relation. Prior to joining Relation, he was Chair of Scientific Translation at the Francis Crick Institute, where during his eight-year tenure he held several leadership positions. Previously, Dr. Roblin was SVP and Head of European R&D for Pfizer. He is also a non-executive director of Sosei Heptares (SOLTF) and non-executive chairman of Centauri Therapeutics.
Dr. Lindsay Edwards was previously Vice President and Head of Artificial Intelligence for Respiration and Immunology at AstraZeneca, where he implemented a world-class machine learning capability. Previously, he founded and led one of GSK’s early data science groups and was Vice President and Head of Artificial Intelligence/Machine Learning for the UK and Europe.
Relation’s Scientific Advisory Board includes luminaries from a wide range of disciplines, including Professor Michael Bronstein, DeepMind Professor of AI at Oxford University and Head of Graph AI at Twitter; Professor Caroline Uhler, Principal Fellow of the Institute and Co-Director of the Eric & Wendy Schmidt Center at the Broad Institute and Associate Professor at MIT; and Professor Alex K. Shalek, Institute Fellow at the Broad Institute and Associate Professor at MIT. Relation has a history of collaboration and co-editing with Turing Award-winning Professor Yoshua Bengio of the Montreal Institute for Learning Algorithms (Mila) on the application of active learning algorithms.
In addition to DCVC and Magnetic Ventures, the seed funding also included participation from Khosla Ventures, OMERS Ventures and firstminute Capital, Peer Schatz, former CEO of Qiagen (NYSE:QGEN), Jonathan Milner, founder of Abcam (LSE:ABC ), and Mark Stevenson, former COO of ThermoFisher (NYSE: TMO).
Relationship discovers new biology to cure disease faster, transforming the traditional R&D model: because patients wait. An integral part of this, Relation combines the power of active graph machine learning and large-scale, high-quality data that describes disease biology to find better ways to target human pathology. We are generating proprietary maps of diseases in humans, leveraging human genetics, single-cell omics, and perturbing data, so that these powerful models can finally be deployed. The Relation platform is in an integrated loop of wet lab, dry lab and translational science, located in London.