Recursion

The promise of TechBio. Investment firm DCVC has released its Deep Tech Opportunities Report highlighting the areas where they see the most promise and potential, from robotics and quantum computing to defense, TechBio, TechMed and drug discovery. ▪ The report notes that when it comes to TechBio, which they define as ““a new branch of the AI industry that constructs proprietary biological and chemical datasets to fuel advanced computation and raise the hit rate in drug discovery,” the biggest bottleneck has been the complexity of biology itself. ▪ But by gathering data in a systematic, automated way and using in silico modeling, they write that companies can now make increasingly precise predictions, much faster and cheaper than traditional drug discovery. ▪ The report notes that Recursion “started doing TechBio long before there was a common term for it” and focused efforts on amassing the largest collection of biological and chemical fit-for-purpose datasets in order to drive new discoveries – a collection that has now grown to over 50 petabytes of data – “including phenomics data showing how toxins, pathogens, genetic changes and candidate drug molecules affect the morphology of individual cells.” ▪ “They have one of the biggest data sets out there, which puts them at the vanguard of the data race in biopharma” says Zachary Bogue. 👉 Read the full report here: https://lnkd.in/e76eUvVe #ai #deeptech #techbio #opportunities #data

🎉 Congratulations to Marissa Saunders, Senior Director of Data Science, who won the Data Innovator Award at the Women Tech Council’s Women Tech Awards last night! 🎉 ▪ Marissa has over 15 years experience in data science and biosciences, and leads teams at the interface of deep learning and drug discovery who are designing and arranging data science and experimental modules into seamless end-to-end processes that enable us to make unbiased decisions about which therapeutic programs to advance. ▪ She received her PhD in Chemistry from the University of Chicago, and did a post doc at the University of Utah in a structural biology lab, and has earned multiple patents for her work.  ▪ Beyond the lab, Marissa is a fierce advocate for diversity in STEM and once competed in the US Open Muay Thai competition. Congrats to Marissa and the other winners! #women #womentech #award #tech #ai #drugdiscovery https://lnkd.in/gTvD3Wb

Looking forward to Najat Khan, PhD's keynote at the inaugural "State of AI in Drug Discovery" virtual event from Genetic Engineering & Biotechnology News on Oct. 30!

The story of REC-1245. REC-1245 is our first program to reach clinical studies that derived from our AI-enabled Maps of Biology. To build these maps, we treat various human cell types with a range of biological perturbations such as CRISPR-based knockouts, cytokines and small molecules, to generate high throughput microscopy images that are designed to be readable by machine learning. The model converts these images into numerical representations and can identify promising relationships between genes and compounds, across trillions of possibilities. Using this approach, we identified RMB39 as a novel target which looks functionally similar to the well-known but hard to drug cancer target CDK12.  We then validated the initial hypothesis via iterative cycles of structure-activity relationship, to optimize and create new chemical entities (NCEs). We onboarded these NCEs back into our maps to evaluate the effect in cells against our desired target. Using this method, we developed several small molecule drug candidates and progressed the most promising one – REC-1245 – from hit to preclinical drug candidate in under 18 months – more than twice the speed of industry average. Learn more: https://lnkd.in/gkX2TzzT #ai #ml #cancer #target #drugdiscovery #pipeline #clinicaltrials

*New Chief People & Impact Officer alert!* We are thrilled to announce that Erica Fox has joined us as Chief People and Impact Officer where she will help deliver business goals while focusing on organizational productivity, engagement and growth, as well as Recursion’s broader social impact. Most recently, Erica was Chief People Officer at the AI/ML tech firm Primer.ai, and she previously spent 15 years at Google leading various strategic people initiatives including global onboarding, learning and development, annual strategic and operational planning, and resource management for the global People Operations. A self-described “systems thinker,” she’s excited to contribute to Recursion’s mission to decode biology to radically improve lives and to help foster opportunities for Recursionauts to move fast, implement ideas, collaborate and innovate across functions. She’s also looking forward to guiding the Recursion Foundation, our non-profit entity which is helping build sustainable, diverse and equitable life science and technology hubs, beginning with Altitude Lab in Salt Lake City. Welcome Erica! #people #culture #impact #collaboration

San Francisco, we’re coming for you in November! Altitude Lab is taking 15-20 BioTech startups from Utah and underserved regions on the road, connecting them with distributors, biopharma partners, and strategic investors. This is the first stop of Altitude Lab’s commercialization roadshow program, Horizon. Join us on November 7th for a pivotal event co-hosted with the Healthcare Businesswomen’s Association, featuring a panel discussion on "BioTech Frontier: AI and Emerging Startups." We're thrilled to have Neda Razavi, MBA from iSono Health, Marc Martin Casas from Illumina Ventures, and Shafique Virani, MD FRCS from Noetik sharing their insights on the future of BioTech and AI innovation. Along with the panel discussion, the event will offer exclusive networking opportunities with top industry leaders, providing a chance to connect with potential partners, investors, and mentors. The poster session will highlight rising BioTech founders from Altitude Lab's portfolio, showcasing breakthrough innovations in healthcare and biopharma. Registration is now open—grab your spot before it fills up! https://lnkd.in/dRaZU6ic #bayareabiotech #sanfranciscobiotech #biotech #californiabiotech #utahbiotech  #healthcaretech #lifesciencestartup #research #hbasanfrancisco

“I think what’s really exciting about this particular program of Recursion is that this small molecule and novel target came out from essentially a Google-search equivalent, from this giant map of biology that we’ve already built.” – Chris Gibson A story from Christina Cheddar Berk in CNBC highlights REC-1245, the first program from our AI-enabled Maps of Biology to receive FDA IND clearance for Phase 1/2 clinical trials – and which progressed from target identification to IND clearance in less than 18 months, more than twice the speed of industry average. She writes: “Recursion was able to use its artificial intelligence-enabled drug discovery platform to identify an area of biology to target for the treatment of solid tumors and lymphoma, match it with a drug candidate and move all the way to gaining regulatory approval to begin studies in less than 18 months.” “We think that’s a really exciting proof point, not only for us as a company, but I think for the techbio industry as well,” Chris said. 👉 Read more: https://lnkd.in/er77crJv #ai #ml #techbio #drugdiscovery #cancer #biology #maps #ind

Announcing FDA IND clearance for a Phase I/2 clinical trial of REC-1245 – Recursion’s first program to reach clinical studies that derived from our AI-enabled Maps of Biology, resulting in target identification to IND enabling studies in under 18 months. ◾ REC-1245 is a potentially first-in-class, potent and selective RBM39 degrader for the treatment of biomarker enriched solid tumors and lymphoma. ◾ Preclinical data supports that RBM39 degradation induces splicing defects which downregulate DNA Damage Response (DDR) networks and cell cycle checkpoints. ◾ The Recursion AI-Enabled Breakthrough to discover RBM39: Using our AI-enabled Maps of Biology – which allow us to screen millions of predicted biological and chemical relationships – we identified RMB39 as a novel target which looks functionally similar to the well-known but hard to drug target CDK12 ◾ Key Differences in RBM39 vs CDK12  CDK12 is a well-known, yet challenging, cancer target. Many investigational drugs designed to target CDK12 unfortunately also inhibit structurally similar CDK13, resulting in significant toxicity. We leveraged the Recursion OS, exploring predicted relationships across the entire genome and chemical compounds to identify the novel target relationship of RBM39, which our maps suggest has similar biological function to CDK12. Using our platform, we then identified and optimized small molecules, including our candidate molecule REC-1245, that targets RBM39 and has similar effects to loss of CDK12 without affecting CDK13. ◾ Significantly Faster than Industry: We progressed from target discovery to IND enabling studies (the step before potential clearance to enter the clinic) in under 18 months – more than twice as fast as the industry average. ◾ The Phase 1/2 clinical trial will evaluate the safety, tolerability, pharmacokinetics, pharmacodynamics, and potential monotherapy efficacy of REC-1245, and is expected to initiate in Q4 2024. We estimate that the initially addressable population for this potential therapeutic will be >100,000 patients in the US and EU5. Learn more: https://lnkd.in/gkX2TzzT Watch our video about REC-1245 here: https://lnkd.in/gzzMGgHS #techbio #biotech #ind #ai #ml #drugdiscovery #target #cancer #cdk12 #rbm39 #automation

Unlocking a new area of application for ML in cell biology. In a new preprint, researchers from Valence Labs, Recursion’s research arm, reveal MolPhenix, a foundation model that can predict the effect of any given molecule and concentration pair on phenotypic cell assays and cell morphology. ◾ While multimodal approaches such as Dall-E, Sora, and Gemini have unlocked incredible new capabilities in NLP and computer vision, applying these approaches to molecular biology has been more challenging. Typically, multimodal representation learning techniques rely on paired samples to learn common representations, but paired samples are difficult to collect in biology where measurement devices often destroy the samples. (see here: https://lnkd.in/evn_BzBp) ◾ To train a model with biological data, one method is to link specific molecular perturbations with corresponding high-throughput morphology experiments. But there are still issues to overcome: 1) generalization across microscopy images remains challenging;  2) most molecules are inactive and will not produce any distinguishable effect;  3) concentration changes the outcome. ◾ MolPhenix, a molecular phenomics model, addresses these three challenges by: 1) Pre-training using foundation models like Phenom-1, simplifying the images into vectors, removing batch effects by normalizing with controls, and further reducing noise levels by averaging multiple replicates of the same experiments. 2) Leveraging the activity landscape with the help of a foundation phenomics model and the proposed S2L loss and defining an activity landscape to help over-sample the active molecules and adapt the loss function. These training modifications allowed them to consider pairs to be negative if their activity landscape is different. 3) Conditioning the loss and the molecular model on the dose. 👉 Read the paper here: https://lnkd.in/ekeMTSzj And check out the related article! 👇 Congrats to the team! Philip Fradkin, Puria Azadi Moghadam, Karush Suri, Frederik Wenkel, Ali Bashashati, Maciej Sypetkowski, Dominique Beaini #ai #ml #foundationmodel #molphenix #molecules #tech #techbio #drugdiscovery #biology

Today we are honoring Rare Cancer Day. A core Recursion value is "We Care," and that includes support for our patient communities and those impacted by the diseases in which we work. In support of our oncology programs and cancer patients we are beginning new patient advocacy partnerships with the Rare Cancer Coalition and the American Cancer Society. Learn more about rare cancers at National Organization for Rare Disorders: https://lnkd.in/eh2gwb8J #cancer #RareCancerDay #rarecancer