Recursion’s Post

“This is a once-in-a-lifetime opportunity.” In a Fireside Chat with Recursion co-founder and CEO Chris Gibson at Download Day, NVIDIA founder and CEO Jensen Huang shared his vision for the future of healthcare, and how the companies are working together on a data-driven approach to drug discovery that has the potential to radically transform medicine.  ◾ Why he’s excited about this collaboration: “If you look at the foundation of Recursion, three things are in play – the invention of a new algorithm, a new family of algorithms that we call deep learning; supercomputing capabilities that you’ve used here and that we’ve partnered with you to create; and the know-how of processing and extracting the meaning of biology that’s embedded within life, within that data.” ◾ With accelerated computing and generative AI, we are changing “the type of problems we can solve,” he added. ◾ NVIDIA has been an incredible partner as we push to develop the next generation of medicines for patients in need -- including their support in building BioHive-2, the largest supercomputer in the pharma industry, which allows us to run more experiments, build more data, and develop even more powerful foundation models. Watch the full Fireside Chat here: https://lnkd.in/evqAm3QB The full day of presentations and slides from Download Day, featuring experts from across the Recursion ecosystem sharing the latest developments on our platform, pipeline and partnerships, is available here: https://lnkd.in/e9bRKvqf  #techbio #future #ai #health #medicine #pharma #tech

Announcing the world’s first neuroscience phenomap – “Neuromap” – which has been optioned under Recursion’s collaboration with Roche and Genentech. It’s the first of several neuroscience phenomaps possible under our partnership agreement. ◾ The optioning of the neuroscience phenomap triggers a $30 million milestone payment to Recursion. ◾ The neuromap was designed using purpose-built neuronal data, computer vision, and advanced AI algorithms to accelerate new insights and treatments. ◾ To build it, we produced over 1 trillion hiPSC-derived neuronal cells using our advanced cell manufacturing technology, making us one of the most prolific producers of hiPSC-derived neuronal cells in the world. ◾ We collaborated with Genentech to adapt a hiPSC-derived neuronal model and protocol for the creation of a unique phenomap that can be utilized to uncover novel insights in neuronal biology. ◾ In addition, we developed and scaled novel genetic perturbation techniques.  We’re combining this huge phenomic dataset with Genentech’s proprietary transcriptomics data and neuroscience expertise to build multi modal maps "Acceptance of this neuroscience phenomap underscores Recursion’s pioneering work driving into some of the most complex cellular contexts in neuronal science. With this map, and together with our partners, we now can navigate and evaluate potentially novel neuroscience targets at a breadth and scale that is unprecedented,” said Chris Gibson, Ph.D., Co-Founder and CEO. ”We look forward to initiating new programs and building more maps on the forefront of the field with our colleagues at Roche and Genentech.” Neurological conditions are a leading cause of poor health and disability worldwide and have posed unique challenges for scientists due to lack of genetic patient data, poor understanding of underlying pathologies, and the difficulty of designing drugs that can safely cross the blood-brain barrier. #ai #ml #techbio #pharma #tech #drugdiscovery #neuro #neuroscience #neurological #phenomics #map

Recursion just announced it will report second quarter 2024 business updates and financial results on August 8th.  In addition, tune in at 5pm ET/3pm MT on Thursday, August 8, 2024 for our L(earnings) Call – where we provide a real-time livestream of the Company’s latest business developments from Co-founder and CEO Chris Gibson. Submit your questions for consideration here: https://lnkd.in/gAFrxggW   Where to watch: LinkedIn: https://lnkd.in/gM4r28fd  Twitter: https://lnkd.in/gn52iH5F  YouTube: https://lnkd.in/gNx66QJK #techbio #biotech #earnings #RXRX

Why do we need multiple models? Join us at Ai4 - Artificial Intelligence Conferences in Las Vegas Aug. 12-14, where Lina Nilsson, SVP of Emerging Technologies, will give a talk on “Harnessing AI: Transitioning from Solo Models to Integrated Data-AI Loops.” ◾ “Sustained business success via AI is rarely about any single model, algorithm or AI tool,” she says. “Instead, AI success requires applying the right models, based on the right data, to the right problems -- in a way that is integrated with other processes across the business to create measurable impact.” ◾ She’ll provide lessons and strategic insights from Recursion's 10 years as a TechBio leader, using examples and real data from across themes of large-scale compute, massive de novo data collection, model approach optimization for specific business challenges, and rigorous performance and ROI tracking. ◾ And she’ll demonstrate how we have applied iterative loops of data generation and model building for ever-increasing business impact. Learn more & register: https://ai4.io/vegas/ #ai4 #tech #ml #conference #drugdiscovery #techbio #biotech #pharma #data 

Learn how Najat Khan, PhD, Chief R&D Officer and Chief Commercial Officer, is leading efforts to accelerate the pace of data science and AI innovation and adoption across pharma through a roundtable called DISRUPT-DS.

Stitching together data to shift the pace and scale of discovery. In his welcome at Download Day, Chris Gibson explained how we are moving from point solutions to an end-to-end workflow that will drive the TechBio transformation of drug discovery. “We believe that the companies that can stitch together all of these different point solutions into a workflow are the companies that are going to recognize the compound efficiencies of technology across discovery, translation, and development,” he told the audience. He pointed to a number of ways we are not just building the technology, but the ways to interface with that technology – and training foundation models that work across multiple data layers. “No one dataset is going to be sufficient to truly decode biology or chemistry or radically improve lives at scale,” he said. “But by stitching together one of the most compelling sets of data, across multiple different levels – phenomics, transcriptomics, proteomics, inVivomics, genomics, and patient data – and training causal AI models across all of these different layers, we believe that we will be able to start to extract extraordinary signatures of biology not just at the gene or protein level, but at the organismal level.” That end-to-end workflow, he said, will bring a “fundamental shift in pace and scale of discovery.” #techbio #pharma #data #biology #drugdiscovery #discovery

Highlights from the recent Molecular Machine Learning Conference hosted by Valence Labs. "Topics like protein binding, protein folding are interesting problems that have seen tremendous advances in the field," says Dominique Beaini. "How the industry applies these models, and how to properly evaluate the models that are being developed.” The event featured over 45 posters from researchers in industry and academia and more than 180 attendees. The next MoML is coming Nov. 5. Learn more and register here: https://www.moml.mit.edu/ #ai #ml #conference #drugdiscovery Mila - Quebec Artificial Intelligence Institute Jonathan Hsu

From zero to supercomputer. ◾ A couple months ago, we announced BioHive-2, the 35th most powerful supercomputer in the world and the top supercomputer in the pharma industry. ◾ Our team, along with a team from NVIDIA, worked over the course of three intense weeks to bring BioHive-2 to life. ◾ Biohive-2  is four times faster than our original supercomputer, BioHive-1, and allows us to perform multiple ambitious projects with shorter timeframes, unlocking even more insights from our 50+ petabytes of proprietary data. ◾ With it, we are able to train larger, more generalizable foundation models in order to accelerate the industrialization of drug discovery. Here’s a time lapse video of the supercomputer build…and wait for the special guest appearance! #ai #supercomputer #techbio #biotech #pharma #drugdiscovery #paintingwithbob

Pushing the boundaries of what’s possible in automation. Paul Rearden spent 15 years in pharma as an ADME scientist. Now he’s leading a group of diverse scientists working in in vivo pharmacology, bioanalytical chemistry, DMPK, and discovery pharmaceutics – and helping to push into the boundaries of what’s possible in automation. Here, he shares our approach to automated drug discovery -- and why this is a pivotal moment for scientists. 1️⃣ Talk about Recursion’s high throughput in vitro ADME platform for early compound screening. In order to create a truly automated lab, we needed to streamline data generation and experiments. Working across a team of software engineers, data scientists, biologists, chemists and technicians, we have built a state-of-the-art automated wet lab that is designed for training machine learning models. As the high quality data grows, the models improve, in a continuous virtuous loop. We needed several essential elements to build this lab, including a single assay, carefully controlled in a homogenous environment with well-defined optimized parameters. We’ve implemented high throughput, LC-HRMS analysis, with sophisticated error recovery systems that minimize human input and instrument downtime. Our platform can be monitored remotely with webcams and real time data status readouts. Processing the large volume and breadth of data has similarly been reduced to confirming QC acceptance.   We are constantly scaling our capacity and improving our data generation and models. Currently our automated lab performs 90x the throughput of manual labs, and tests over 750 compounds per week in a range of assays. 2️⃣ What is the value of automation? The earlier you can de-risk and throw out bad molecules, the more time and money you save. You take critical predictors of future in vivo success and automate it. Over multiple experiments on stability, binding, and permeability, we generate results that we can predict. With our AI and ML colleagues and our industry leading supercomputer, we want to run these models on everything -- deploy our richer datasets, and we’ll outperform other approaches. 3️⃣ Talk about the role of human scientists. With increased automation, we’re freeing human scientists to design the next thing. We’ve learned a lot – it’s harder than we thought it was going to be to run at this scale but the team is  progressively moving toward more and more autonomy. We’re building predictive models from this data utilizing Recursion’s cutting edge expertise and compute. This is an important moment for the careers of these scientists – they understand it’s about the bigger picture. We’re going to build the next generation of our field, marrying big data and predictive approaches with classical understanding of the underlying science we’ve built upon. #ai #automation #techbio #science #tech #biology #pharma #drugdiscovery