The preprint and corresponding datasets give a glimpse into early progress to industrialize drug discovery by decoding biology
SALT LAKE CITY–(BUSINESS WIRE)–Recursion, a digital biology company industrializing drug discovery, released a preprint on applying deep-learning-driven analysis of cellular morphology to develop a scalable “phenomics” platform. The preprint demonstrates the capabilities of Recursion’s platform to model complex immune biology and screen for new therapeutics.
Despite incredible efforts by dedicated scientists around the globe, new drug approvals still come at a rate of many dozens per year at best, and many of these approvals are for drugs that improve on prior treatment. At the same time, thousands of named diseases today have no approved treatment, and due to its complex and interdependent nature, drug discovery in the immune system is a particularly challenging endeavor.
In these proof-of-concept studies, Recursion demonstrates elements of its deep-learning enabled phenomics platform to functionally map a massive network of cellular immunity, employing a vast panel of immune stimuli across key cellular contexts. These high dimensional representations of complex immune behavior enabled Recursion to conduct high-throughput screens of nearly 100,000 potential drug starting points, and to predict the target of key candidates to enable rapid prioritization or deprioritization for further work.
“I am excited to share this tiny sliver of our work with the world,” said Chris Gibson, Ph.D., co-founder and CEO, Recursion. “We have the ambition to build a machine that will industrialize drug discovery through the application of technology at each step in the discovery process. Our ability to simultaneously advance dozens of programs, including four at the clinical stage, with a relatively small team, is a testament to the disruptive power of technology thoughtfully applied. And we are just getting started — with new efforts in digital chemistry, predictive ADMET and with our recent acquisition of Vium, the digitization of the vivarium. As we assemble all these components and make them work together seamlessly, we believe we can pioneer the next generation of digitally native biotech companies.”
This paper shares preliminary results from approximately 220,000 unique combinations of compounds, cells and disease perturbations Recursion has conducted in the past 12 months using a fraction of its bandwidth. Beyond the immunology applications shared today, Recursion has conducted and analyzed a further 33 million experiments on its in-house automated wet-lab platform spanning rare disease biology, fibrosis, immunology, inflammation, neuroscience, infectious disease, aging and senescence, oncology and immuno-oncology. By generating these ever-growing phenomics data alongside algorithms enabling their relatability across experiments, time and disease modality, Recursion is building functional maps of human biology. These datasets will be continuously integrated to allow for unbiased drug and target discovery.
Continuing with its commitment to open-source some of its datasets, Recursion is depositing the images and embeddings for findings in the preprint at RxRx.ai in hopes that they can provide insights that enable researchers to explore and uncover insights from the data. Combined with Recursion’s RxRx1 dataset released last year, RxRx2 and RxRx19 released today enable machine learning researchers to leverage modern deep learning techniques to study complex immune phenomena. To download the free RxRx datasets, visit https://RxRX.ai/. For more information on Recursion’s unique approach to applying artificial intelligence and machine learning to drug discovery and development, visit www.recursionpharma.com.
Recursion is a digital biology company industrializing drug discovery. Recursion does this by combining automation, artificial intelligence, machine learning, in vivo validation capabilities and a highly cross-functional team to discover novel medicines that expand our collective understanding of biology. Recursion’s rich, relatable database of over 5 petabytes of biological images generated in-house on the company’s robotics platform enables advanced machine learning approaches to reveal drug candidates, mechanisms of action, novel chemistry, and potential toxicity, with the eventual goal of decoding biology and advancing new therapeutics that radically improve people’s lives. Recursion is proudly headquartered in Salt Lake City. Learn more at www.recursionpharma.com, or connect on Twitter and LinkedIn.