Weekly TechBio News (part I)
MetaphysicalCells: A newsletter about Science, Technology and AI Drug Discovery
Welcome back to another free edition of MetaphysicalCells 🪐🛸
Below is a synthesis of key trends and insights from the latest TechBio updates:
➡️ Weekly TechBio News (part I).
Weekly TechBio News (part I)
💈 Araceli Biosciences Inc
Portland-based biotech company Araceli Biosciences raised on April 24, 2025 $11.2M in new funding to fuel the expansion of its high-content imaging 🔬 technology and AI-powered analysis (Araceli Biosciences raises $11.2M for imaging and AI analysis systems to help speed drug discovery).
Araceli Biosciences founded in 2022 in the US is utilizing AI for high content analysis during drug screening. Right know they are offering 3️⃣ solutions for diverse applications:
Araceli Endeavor® for complete scan in under 10 minutes (full well, full plate) in 4 channels at high resolution (0.27 micron per pixel, with full-well image fidelity).
Araceli EndeavorGT™ delivers 6,000 wells per hour, 35 plates every working day and 875 plates every month.
Clairvoyance™ SOFTWARE delivers Neural Net Optimized Segmentation, utilizing existing trained neural networks for segmentation of cells, spots, nuclei, networks, etc. Unlike classical image analysis which relies on processing and intensity based segmentation, Clairvoyance uses dedicated neural networks to segment features to be analyzed. DL-powered neural networks are capable of handling differences between cells, wells, batches and even assays.
On January 23, 2025, VALID Therapeutics, a Techbio startup dedicated to reshaping drug discovery with complex human induced pluripotent stem cell (iPSC) co-culture models that deliver increased predictive validity, partnered with Araceli Biosciences to leverage the cutting-edge Araceli Endeavor® Ultra High Throughput - High Content Imaging Platform (VALID Therapeutics Partners With Araceli Biosciences to Advance AI-Driven Drug Discovery).
🆕 Kids on the block: Grantease & ScienceMachine
💈 Grantease IO
Grantease AI-Powered Grant Writing Platform is an innovative, AI-powered solution that transforms the grant writing process. The platform streamlines the complex, time-consuming journey of securing grant funding, making it easier and more effective for organizations across sectors—healthcare, housing, scientific research, religious institutions, and more—to access the resources they need.
In just 15 minutes, Grantease's specialized AI generates:
✅ An up-to-date literature review: Automated literature reviews with access to over 360 million papers.
✅ Detailed and styled Specific Aims, Abstract, Narrative & Research Strategy and references in minutes.
✅ First draft with inline citations, ready for review and revisions.
As a matter of fact, with Grantease you can 📍Unlock a specialized AI model tailored to your RFA (Access AI trained for your NIH grant type, Identify all compliance requirements instantly, Pinpoint key funding priorities); 📍Generate Your Lit Review (Search 360+ million research papers instantly, Get the most relevant citations, Build a compelling evidence foundation); and 📍Review Your First Draft (See your proposal written in your unique voice, Verify all claims with reliable citations, Maximize your chances of funding with NIH specific editing tools).
🗣️ Top 18 Online AI Courses for Grant Writers
Supercharge your grant writing with AI! Discover top online courses that teach grant writers of all levels to use AI tools effectively and ethically.
💈 ScienceMachine AI
ScienceMachine’s goal is to accelerate biomedical discovery, by developing AI systems that help scientists put large parts of their research on autopilot, starting from data analysis. For this reason they created Sam, the first AI Bioinformatician, which structures, analyzes, and uncovers new insights in your research data, dramatically enhancing your R&D capabilities. Sam builds and maintains customized data workflows for your company, quickly learns about your specific scientific needs, and works around the clock to support your research.
For example, you can ask
🗣️ “Hey Sam, could you please check if patient demographics (e.g. age, gender, BMI, etc.) have a significant impact on the likelihood of remission (CR/PR) with camrelizumab in this HCC cohort?”
🗣️ “Hey Sam, I've attached some data from a GWAS study in ALS. Could you please check what pathways the affected genes enrich for? In particular, I am interested if any of the pathways are involved in glycine metabolism.”
🗣️ “We performed RNA-seq of depressed mice, some of which received Paroxetine. Please check if treatment with Paroxetine leads to any tissue-specific changes in gene expression.”
Sam works 24/7 to discover new insights in your organisation's research data, allowing you to improve experiments, optimize bio-products and find new opportunities. They tailored Sam to your specific science and research needs. Basically they integrate their AI within your company and fine-tune it to your specific research needs. Then experts double-check all outputs generated by the AI, making sure you get accurate results and can achieve your R&D goals. Once the AI is tailored and integrated, you can keep using it on your own! The full version of Sam integrates with your existing tools and data sources to make sense of large, complex R&D data—helping improve experiments, optimize bio-products, and find new opportunities.
Any data they analyze is:
📍 Never used for training a model, you own your inputs and outputs, they just provide the AI tool.
📍 Encrypted with enterprise-grade security practices, trusted by dozens of life sciences companies.
📍 Always treated privately and never shared with third parties. Only you have access to your data, and your computation is isolated with no contamination.
Lorenzo Sani and Benjamin Tenmann are co-founders at ScienceMachine.
💈 Tempus AI Inc, Pathos AI and AZ
On April 24, 2025, AstraZeneca, Tempus and Pathos AI signed a multi-year agreement to develop a large-scale multimodal DL model (AstraZeneca enters $200m AI cancer pact with Tempus and Pathos) that will be used to extract biological and clinical insights, identify novel drug targets, and support the development of new cancer therapies. Under the terms of the deal, Tempus will receive $200M in data licensing and model development fees, and will contribute its large library of de-identified oncology data to help build the model, which will be shared by all three parties once completed.
🔹Tempus🔹
Tempus AI Stock Sees 21% Gain in 3 Days: Is it a Buy After AZN Deal?
Why Tempus AI Stock Is Skyrocketing Today (April 23, 2025)
The Chicago-based Tempus—that specializes in AI and precision medicine and has one of the world’s largest libraries of clinical and molecular data—provides integrated solutions such as sequencing, companion diagnostics, clinical trial solutions, data collaborations, biological modeling, and etc by applying AI. Apart the three companion diagnostic solutions (xT, xR, xF ), Tempus is also offering omics solutions, biological modeling (PanKRAS, PanCRC), sequencing (Tempus xT CDx*, Tempus xT, Tempus xR, Tempus xF+, Tempus xE, Tempus xG/xG+, Tempus xM Monitor and Tempus xM MRD) and data collaborations.
Almost a month ago (April 11, 2025) Tempus AI (NASDAQ: TEM) announced it has acquired Deep 6 AI, a leading AI-powered precision research platform for healthcare organizations and life sciences companies (Tempus Announces Acquisition of Deep 6 AI). Deep 6 AI, utilizes AI to mine medical records to accelerate finding and recruit patients for clinical trials within minutes. Its AI-powered software matches patients to clinical trials by mining real-time structured and unstructured electronic medical record (EMR) data across a broad ecosystem, which includes academic medical centers, National Cancer Institute (NCI)-Designated Cancer Centers, and NCI Community Oncology Research Programs. Deep 6 is integrated with over 750 provider site locations spanning more than 30 million patients, which will materially add to Tempus’ existing network.
Further to this acquisition, Genetics Corp. (a medical diagnostics and testing company) announced on November 4, 2024 an agreement to be acquired for $600M by Tempus AI Inc (Tempus AI to Buy Ambry Genetics for $600M).
Apart the last collaboration with AZ, Tempus has recently announced also the following partnerships ➡️
Tempus and🔺 Northwestern Medicine Announce Collaboration to Bring Practical Application of AI to Healthcare,
Tempus Announces Collaboration with🔺 IFLI Aimed at Supporting Development of Targeted Therapies for Follicular Lymphoma,
🔺 Artera Announces Collaboration with Tempus to Expand Access to Personalized Prostate Cancer Treatment,
🔺 Imagene Announces a Collaboration With Tempus to Advance AI-Driven Diagnostics for Non-Small Cell Lung Cancer,
🔺 Avacta partners Tempus AI to boost cancer therapy development,
🔺 SoftBank partners with Tempus AI to launch AI healthcare venture,
Tempus Announces Collaboration With🔺 JW Pharmaceutical to Apply Real-World Data and Biological Modeling to Enhance Early Research and Development.
Finally, just this week (April 28, 2025) Tempus Introduced Loop, an AI-Powered Target Discovery and Validation Platform. Loop is Tempus’ proprietary approach to novel target identification that integrates real-world patient data (RWD) with human-derived biological models and CRISPR-screens, all leveraging AI to rapidly uncover insights for pre-clinical therapeutic development.
Few words about Pathos and AZ now.
Pathos is a clinical-stage biotechnology company focused on re-engineering drug development, leveraging the power of advanced AI technologies to bring precision medicines to market through partnerships with biopharmaceutical companies. The company was founded by Lefkofsky, who is CEO of Chicago biotech firm Tempus, as well as Fukushima, who holds the chief operating officer role at Tempus. PathOS Platform™ is fueled by petabytes of real-world oncology data and patient-derived functional genomic data to identify and validate compelling biology that would have otherwise gone undetected using traditional methods. Currently they are working on 2️⃣ assets in their pipeline:
Pocenbrodib (previously FT-7051) in Phase 1 is a small molecule inhibitor that has the potential to provide clinical benefit for patients with advanced prostate cancer, either alone or in combination with other treatments (Pathos AI Doses First Patient in Phase 1b/2a Clinical Trial of Pocenbrodib, a CBP/p300 Inhibitor, April 2025).
P-500 (previously PRT811) in Phase 2 is a selective, brain-penetrant small molecule inhibitor of protein arginine methyltransferase 5 (PRMT5) that has the potential to provide clinical benefit for patients with advanced solid tumors, including high-grade glioma and uveal melanoma.
AstraZeneca plc (AZ), the British-Swedish multinational pharmaceutical and biotechnology company with its headquarters at the Cambridge Biomedical Campus in Cambridge, England, is involved in the following AI drug discovery collaborations:
🖇️ In October 2024, Owkin and AstraZeneca Partnered on AI-Powered gBRCA Screening for Breast Cancer.
🖇️ In September 2024, Immunai an Israeli AI startup forged $18m partnership with AstraZeneca for cancer drug trials.
🖇️ In May 2024, Nabla Bio Secured $26M Series A Financing and Collaborations with AstraZeneca, Bristol Myers Squibb and Takeda for Generative Protein Design.
🖇️ On December 06, 2023, AbSci announced a potential $247M partnership with AZ to focus on the discovery of novel cancer treatments with the help of AI technology.
🖇️ On October 23, 2023, Nanoform announced that it had granted AZ a global online STARMAP® license, a digital AI version of the CESS® technology that enables in-silico experiments to determine which molecules should be nanoformed.
🖇️ In September 2023, AZ's rare disease unit Alexion and AI-drug discovery biotech Verge Genomics (one of industry's most advanced all-in-human AI powered drug discovery companies) have partnered to identify new drug targets for rare neurodegenerative and neuromuscular diseases (AstraZeneca converges with Verge's AI tech in $42M neurodegenerative disease R&D pact).
🖇️ For previous to 2023 collaborations: AI Drug Discovery News (3): AstraZeneca and Novartis.
💈 Turbine AI (Turbine Simulated Cell Technologies Ltd)
On April 24, 2025, Champions Oncology (NASDAQ:CSBR) (Champions Oncology: Global Oncology CRO Services), a global leader in oncology R&D solutions, announced a strategic collaboration with Turbine (Champions Oncology's Data Platform Gains Momentum through a Strategic Collaboration with Turbine), integrating Champions' state-of-the-art comprehensively annotated, clinically relevant multi-omic dataset with Turbine's cutting-edge virtual lab technology and cell-based simulation platform, to conduct in silico experiments covering perturbations they would otherwise be unable to investigate. This significant synergy minimizes reliance on traditional laboratory methods and enhances drug discovery initiatives, ranging from target identification to the generation of mechanistic insights, through the integration of highly translational multi-omics datasets.
Turbine’s in silico experiments can test an almost infinite number of interventions on a simulated cell that accurately reflects the molecular diversity of cancer cells. More precisely, Turbine’s core technology, the Simulated Cell™ platform, utilizes ML to understand how human cells make decisions, by mapping and modeling how thousands of signaling proteins interact, and by characterizing cellular level cancer behavior and response or resistance to treatment. In other words, Turbine takes laboratory trial and error out of drug discovery and tackles it with scalable power on its servers, all by leveraging AI in an evolving, predictive platform to understand cancer mechanisms and unlock novel therapies with Simulated Cells. On May 16, 2024, Accenture (NYSE: ACN) has made a strategic investment, through Accenture Ventures, in Turbine to further extend its capabilities to global biopharma companies that can benefit from Turbine’s ability to uncover hidden biological insights, with the potential to guide and accelerate key drug development workstreams (Accenture Invests in Turbine to Accelerate Use of AI-Powered Cell Simulations for Biopharma Research and Development).
The London based Budapest University spinout Turbine AI was established in 2016 by Dániel Veres, Kristóf Zsolt Szalay, and Szabolcs Nagy Turbine, and has partnerships with several firms, including 🔹Harmonic Discovery, 🔹Bayer (ETR: BAYN), 🔹AstraZeneca (LON: AZN), 🔹Ginkgo Bioworks (NYSE: DNA) and 🔹Ono (TYO: 4528). In particular, on September 12, 2024 Turbine announced that Ono Pharmaceutical Co., Ltd (Osaka, Japan) has selected multiple targets identified using Turbine’s in silico Simulated Cell™ platform for further development (Turbine Achieves Key Milestone in Collaboration with Ono Pharmaceutical). Additionally, they have 3️⃣ internal programs (Genome instability & mutation, Sustaining proliferative signaling, Deregulating cellular metabolism), 1️⃣ novel kinase program NEK1 in collaboration with 🔹Harmonic Discovery, 1️⃣ CDC7 program in collaboration with 🔹Cancer Research Horizons, 3️⃣ programs with 🔹Bayer, 2️⃣ programs with 🔹AstraZeneca and 2️⃣ more programs with other biotech/pharma partners.
💈 Lokavant spun out from Roivant Sciences Ltd
Lokavant has just launched the AI-powered clinical trial developing tool 👩🔧 Spectrum v15, to dispel uncertainty in the feasibility of studies (Lokavant launches Spectrum v15 trial forecasting solution). Powered by GenAI, ML, and Causal AI, Spectrum turns complex feasibility inputs into adaptive forecasts you can trust.
Their solution leverages ML, causal AI, and generative AI models, utilizing a historical dataset from 500,000 studies to predict the likelihood of achieving critical last-patient-in milestones. According to Lokavant, previously, setting up forecasting comparators manually took a minimum of five weeks, but with this solution, this process now takes five minutes.
Spectrum now delivers the following key advantages:
📍 Granular forecasting at all levels: In pre-study and mid-study analysis, study teams can enter site, country, region, and study-level forecasting inputs across enrollment, discontinuation, and screen failure, while accounting for non-enrolling sites.
📍 Adaptive trial design forecasting: Configure forecasts based on adaptive trial designs with the flexibility to define multiple enrollment-driven interim analyses and other study pauses.
📍 Set and achieve screening goals: In pre-study and mid-study analysis, model participant screening volumes at site, country, region, or study level required to achieve study-level enrollment goals based on screen failure, discontinuation, and enrollment rates at each of these geographic levels.
📍 Real-time continuous (re)forecasting: With real-time integration of live study data, study teams can continuously track ongoing study startup and enrollment and generate alternative study scenarios to ensure enrollment success.
Roivant Sciences, Lokavant’s “mamma” (Basel, Switzerland and London and New York) (NASDAQ:ROIV), is building entrepreneurial biotech and healthcare technology companies with a unique approach to sourcing talent, aligning incentives and deploying the technology. So far, they have the following companies in the Roivant family spanning multiple therapeutic areas, modalities and technologies: Dermavant, Immunovant, Priovanttx, Datavant, VantAI, Psivant, Covanttx, Lokavant, Kinevant and Genevant. Moreover, the Roivant Discovery physics-driven drug design platform combines quantum physics, statistical thermodynamics, molecular simulations, AI, ML, a dedicated HPC supercomputing cluster, purpose-built software, and in-house laboratory.
Roivant’s spinout VantAI Inc has recently unveiled Neo-1, an AI model combining biomolecule structure prediction and molecular generation to design new drug candidates, including molecular glues and degraders (VantAI Launched Neo-1, the First AI Model to Rewire Molecular Interactions by Unifying Structure Prediction and Generation for Therapeutic Design).
Neo-1 is the first model to unify de novo molecular generation and atomic-level structure prediction in a single model, by generating latent representations of whole molecules instead of predicting atomic coordinates. Optimized together with NVIDIA, Neo-1 enables design for high-value mechanisms of action, including molecular glues and other proximity-based therapeutics, directly addressing one of pharma’s biggest discovery bottlenecks, and is already in use across three large pharma and biotech partnerships with VantAI.
💈 Thrive Biosciences Inc
Thrive Bioscience Automated imaging and analytics manufactures and sells a family of instruments with software tools that provide previously unavailable live cell imaging, analytics, and automation for improved processes, reproducible results, and breakthrough insights. Thrive Bio’s products leverage imaging and extensive software tools to automatically capture and build a database of terabytes of high-resolution images and data.
The🔺CellAssist Benchtop Imager is an environmentally-controlled, benchtop system that images single cell culture plates from multiple users. The imaging system provides bright-field, phase contrast, and Thrive Phase Imaging at 4x, 10x, and 20x at 100+ focal planes on 6-well through 384-well flat and round bottom plates. The CellAssist Software Suite automatically builds a centralized database of images, metrics, and methods and materials.
The 🔺CellAssist 50 System, High-Throughput Live Cell Analysis Platform, was designed to enable quantitative cell biology. An automated time series cell imaging system with the same imaging and software that drives the CellAssist Benchtop Imager, paired with an integrated 50-plate CO2 incubator, the CellAssist 50 System is ideal for large-scale experiments, or managing multiple projects and users in parallel. With dynamic scheduling, each of the plates can have separate imaging schedules and imaging modes, allowing for consistently timed around-the-clock experiments.
The🔺CellAssist Software Suite, Automated Live Cell Monitoring, Data Management and Advanced Analytics, is the powerful engine behind their innovative image processing and analytics, this software simplifies workflows, enhances data reproducibility, provides robust tools for data management and documentation, combined with a powerful set of metrics to analyze cells, stem cells, and organoids.
💈 Tahoe Bio AI Previously Vevo Therapeutics Inc
Vevo Therapeutics rebranded to Tahoe on April 23, 2025 and established in 2021 in San Francisco, is using its in vivo drug discovery platform and next generation AI models to uncover better drugs for more patients. Vevo is building the world's largest in vivo, single-cell atlas of how chemistry perturbs biology with Mosaic, which is the first platform to generate high-resolution in vivo data with single cell precision while capturing patient diversity in drug response.
In a single in vivo experiment, Mosaic can measure how a drug impacts cells from tens to hundreds of patients, generating millions of datapoints on drug-induced changes in gene expression, by measuring both phenotypic and transcriptomic changes in cell states. By pooling cells from 10s to 100s of diverse patients in one tumor and single-cell RNA profiling, they can determine drug action across all of those patients in one experiment. With this mosaic of patient cells, they can better represent patient diversity in drug response earlier in the discovery process, something that the classical in vitro models usually overlook.
On December 05, 2024, Parse Biosciences announced the generation of the world’s largest single cell dataset to date totaling 100 million cells, created in partnership with Tahoe (Vevo Therapeutics Partners with the Parse Biosciences GigaLab to Generate 100M Cell Atlas for AI Powered Drug Discovery). The new dataset, named Tahoe-100, comprises 100M cells and 60,000 conditions with 1,200 drug treatments across 50 different tumor models. In addition to the AI drug discovery dataset generated with Tahoe, the Parse GigaLab is currently working on large-scale projects for a variety of applications including additional perturbation screens, data for generative AI models and atlasing for population studies.
Parse Biosciences is a global life sciences company empowering researchers to perform single cell sequencing (profiling from 1000 to 1 million cells or nuclei in a single experiment) with unprecedented scale and ease. Their technology Evercode™ split-pool combinatorial barcoding enables you to scale up your single cell projects to millions of cells or nuclei, by labeling 🏷️ molecules with cell-specific combinations of barcodes. Cells are first fixed and permeabilized and then the split-pool barcoding process labels cells with an exponentially large number of barcode combinations, making it possible to easily scale beyond other technologies and up to 1 million cells, also collected on different days and processed in parallel. Moreover, by leveraging Parse’s Evercode technology and expanded automation, the GigaLab enables groundbreaking research with single cell approaches by sequencing 10M cells per run.
In February 2025, Tahoe (Collaborating with NVIDIA on biological foundation models) open-sourced Tahoe-100M, the largest single-cell atlas ever—by a wide margin—and this is a huge leap forward for AI models of cells and drug discovery. Tahoe-100M sets a new bar with:
➡️ 100M single-cell data points (featuring 100 million single-cell data points),
➡️ 60,000 drug-patient interactions (60,000 experiments),
➡️ 50x larger than all public perturbative single-cell data combined, built using their Mosaic platform.
💈 More News 💈
Arc Institute Teams with 10x and Ultima Genomics to Evolve Virtual Cell Atlas
Sonrai and Paige partner to bring advanced machine learning capabilities to biopharma
Making Brain Waves: AI Startup Speeds Disease Research With Lab in the Loop (BrainStorm Therapeutics)
Profluent Introduces ProGen3, Demonstrating Scaling Laws for Foundation Models in Writing Biology
Argonne Leverages AI and Supercomputing to Revolutionize Cancer Research
Immunai & Parker Institute to Build Largest Single-Cell Immune Profiling Patient Database
Flagship Pioneering Unveils Etiome to Pioneer Preemptive Healthcare
Algorae Pharmaceuticals strengthens drug pipeline as AlgoraeOS Version 2.0 development continues