AI For Drug Discovery: Weekly News 🐳
🎐 SK Biosciences partnered with PhnyX Lab
South Korea-based 🇰🇷 SK Biosciences (a biotechnology company that produces vaccines and other products and a subsidiary of SK Chemicals) just signed ✒️ a strategic Memorandum of Understanding (MoU) with AI corporation PhnyX Lab in Silicon Valley (SK Biosciences partners with PhnyX Lab to streamline drug discovery research using AI). Through this agreement, the two companies plan to jointly develop a customized solution that automates tasks such as literature search, data analysis, and document preparation necessary for SK Biopharm's drug development process, based on the generative AI engine Cheiron from PhnyX Lab.
PhnyX Lab, an AI startup established in Silicon Valley in September 2024 by graduates of Stanford University's Department of Computer Science, and backed by the Korean SK Networks (SK Networks Launches AI Lab in Silicon Valley, USA) introduced on December 30, 2024 Cheiron, a generative AI platform for the pharmaceutical industry that leverages cutting-edge technology to simplify drug discovery, clinical trials and production, saving time and resources (Cheiron: PhnyX Lab Startup Unveiled Korea’s First Medicine GenAI Search Platform).
Cheiron has automated the entire process from literature review to report writing by analyzing academic data. It has enhanced accuracy and effectiveness tailored to the pharmaceutical and biopharmaceutical industries by utilizing official databases from major regulatory agencies such as the US Food and Drug Administration (FDA) and the Korean Ministry of Food and Drug Safety, as well as the Medical Subject Heading (MeSH) classification system.
The SK Networks behind PhnyX Lab (KRX: 001740) paving the way for AI, is a subsidiary of the SK Group. The SK Group (Korean: SK그룹; 에스케이그룹) is a South Korean multinational manufacturing and services conglomerate headquartered in Seoul. A chaebol (Korean family-owned conglomerate), the SK Group is the second largest such conglomerate by revenue in South Korea, after the Samsung Group. Through a number of subsidiaries, it is engaged in various businesses, including manufacture of chemicals and petrochemicals, semiconductors, flash memory and miscellaneous information technology, as well providing telecommunications services worldwide among its other less notable ventures.
🎐 Health Abu Dhabi and Sanofi agreed to collaborate, and advance clinical research and leverage mRNA and AI technologies
At the BIO International Convention 2025 in Boston, the Department of Health Abu Dhabi (DoH), namely the healthcare regulator for the Emirate, signed a Memorandum of Understanding (MoU) 🤝 with Sanofi (Next-Generation Vaccines Advance Through Abu Dhabi DoH and Sanofi’s MoU Agreement at US BIO 2025). The agreement outlines a strategic partnership aimed at accelerating the development of next-generation vaccines while reinforcing regional capabilities in vaccine manufacturing. By leveraging Abu Dhabi’s growing health-tech ecosystem and advanced research infrastructure, the collaboration seeks to support global R&D efforts and strengthen supply chain resilience across the Middle East. The collaboration aims to advance clinical research, optimize resources, and leverage mRNA and AI technologies.
🎐 Xaira Therapeutics released X-Atlas/Orion
Xaira Therapeutics, the $1️⃣ billion AI drug discovery unicorn 🦄 co-founded by Nobel laureate David Baker, has released X-Atlas/Orion (Genome-wide Perturb-seq Datasets via a Scalable Fix-Cryopreserve Platform for Training Dose-Dependent Biological Foundation Models), that is an 8️⃣-million-cell open-source dataset detailing not just if a genetic change affects a cell, but by how much (How Xaira aims to fuel biology’s ‘ImageNet moment’ with a 521-GB open-source dataset).
This dataset (X-Atlas/Orion) contains processed data from two genome-wide Perturb-seq experiments in HCT116 and HEK293T cell lines described in the manuscript👉 X-Atlas/Orion: Genome-wide Perturb-seq Datasets via a Scalable Fix-Cryopreserve Platform for Training Dose-Dependent Biological Foundation Models. 👇
🔷 Orion edition (X-Atlas/Orion), is the largest publicly available Perturb-seq atlas generated from two genome-wide FiCS (Fix-Cryopreserve-ScRNAseq) Perturb-seq experiments (capturing perturbation-induced transcriptomic changes) targeting all human protein-coding genes, and comprising 8️⃣ million cells deeply sequenced to over 1️⃣6️⃣,0️⃣0️⃣0️⃣ unique molecular identifiers (UMIs) per cell.
🔷 Xaira’s FiCS Perturb-seq platform, leverages the Chromium platform from 10x Genomics.
🎐 Labviva automates chemical supply with Harper Chemical
Labviva (an AI-powered digital purchasing platform for complete transparency into your supply chain) announced 📢 the general availability of Harper Chemical (HarperChem), a 🆕 part of its cloud-based inventory management system (IMS) (Labviva Automates Chemical Supply Chain for the Life Sciences Industry). Harper Chemical, is an AI-powered chemical search suite that automates supply chain logistics so laboratories can source, store, and replenish vast inventories of raw materials.
On January 13, 2025, Labviva raised $25M to boost AI-powered life sciences procurement platform.
Founded in 2017, Labviva offers an AI-powered digital marketplace that connects researchers with suppliers of reagents, chemicals, consumables and instrumentation, enhancing visibility and insight into the life sciences supply chain.
🎐 A 🆕 cancer drug candidate demonstrated the ability to block tumor growth without triggering a common and debilitating side effect
A 🆕 cancer drug candidate known as BBO-10203 and developed by 👉 Lawrence Livermore National Laboratory (LLNL), 👉 BBOT (BridgeBio Oncology Therapeutics) (a subsidiary of BridgeBio Pharma, Inc, Nasdaq: BBIO), and 👉 the Frederick National Laboratory for Cancer Research (FNLCR) has demonstrated the ability to block ⛔ tumor growth without triggering a common and debilitating side effect (Livermore Lab: Cancer Drug Candidate Developed Using HPC and AI Blocks Tumor Growth).
The discovery of BBO-10203 brings together DOE high-performance computing with AI and biomedical expertise to accelerate drug discovery. LLNL is leveraging its Livermore Computer-Aided Drug Design (LCADD) platform—combining AI and ML with physics-based modeling ⚛️—and world-class DOE supercomputing resources like Ruby and Lassen, to simulate and predict drug behavior long before any compound is synthesized.
In early clinical trials, BBO-10203 has shown promise in disrupting a key interaction between two cancer-driving proteins—RAS and PI3Kα—without causing hyperglycemia (high 🆙⬆️ blood-sugar levels), which has historically hindered similar treatments (BBO-10203 inhibits tumor growth without inducing hyperglycemia by blocking RAS-PI3Kα interaction).
⏩ BBO-10203 is an orally available drug that covalently and specifically binds to the RAS-binding domain of phosphoinositide 3-kinase α (PI3Kα), preventing its activation by HRAS, NRAS, and KRAS. It inhibited PI3Kα activation in tumors with oncogenic mutations in KRAS or PIK3CA, and in tumors with human epidermal growth factor receptor 2 (HER2) amplification or overexpression.
⏩ In preclinical models, BBO-10203 caused significant tumor growth inhibition across multiple tumor types and showed enhanced efficacy in combination with inhibitors of cyclin-dependent kinase 4/6 (CDK4/6), estrogen receptor (ER), HER2 and KRAS-G12C mutant, including in tumors harboring mutations in Kelch-like ECH-associated protein 1 (KEAP1) and Serine/Threonine Kinase 11 (STK11).
⏩ Notably, these antitumor effects occurred without inducing hyperglycemia, as insulin signaling does not depend on RAS-mediated PI3Kα activation to promote glucose uptake.
🐟 Insilico Medicine announced that the 🥇 patient has been dosed in a global 🌐 multicenter clinical trial (NCT06414460) to evaluate ISM3412
Insilico Medicine just announced that the first patient has been dosed in a global multicenter clinical trial (NCT06414460) to evaluate ISM3412, a potentially best-in-class, AI-empowered MAT2A inhibitor with novel structure, in patients with locally advanced and metastatic solid tumors (Insilico Medicine begins clinical trial for AI-designed cancer drug ISM3412).
The Phase 1 study consists of two parts:
🔶 dose escalation and
🔶 dose selection optimization, where participants will receive ISM3412 orally once daily, not only to evaluate the safety, tolerability, pharmacokinetics/pharmacodynamics properties and preliminary anti-tumor efficacy of ISM3412, but also to determine the recommended dose in further studies.
To date, the trial has completed enrollment for the first subject and DLT (Dose-Limiting Toxicities) observation for the first dose cohort at Cancer Hospital Chinese Academy of Medical Sciences, the leading site in China.
🐟 Iambic Therapeutics and Lambda announced a significant partnership
Iambic has selected Lambda (a leading GPU cloud ☁️ provider) to provide a powerful NVIDIA HGX B200 cluster to accelerate the training of its flagship AI model, Enchant (Iambic Therapeutics Taps Lambda's NVIDIA B200 Cloud to Power AI-Driven Drug Development). By leveraging Lambda’s cutting-edge computing infrastructure, Iambic will enhance Enchant’s ability to predict the viability and success of new drug molecules long before they reach clinical trials.
✔️ Iambic.AI (Iambic Therapeutics formerly known as Entos LLC)
Iambic.AI is developing physics-based AI algorithms to drive a high-throughput experimental platform, converting 🆕 molecular designs to 🆕 biological insights each week with: 🔶 NeuralPLexer a protein-ligand structure prediction 3D physics-based equivariant generative diffusion tool, 🔶 OrbNet an AI-accelerated quantum chemistry Graph neural network architecture based on quantum features, 🔶 PropANE a multi-endpoint property prediction tool, and 🔶 Magnet offering generative molecular design.
On March 18, 2024, Iambic stated that its Rapid 🏃 Path to the Clinic Enabled by Its AI-Driven Drug Discovery Platform, was Built in Collaboration with NVIDIA. Then on October 29, 2024, Iambic Therapeutics announced Enchant™, a multi-modal transformer model designed to provide predictive insights into clinical properties of potential medicines from the earliest stages of drug discovery (Iambic Therapeutics Announces “Enchant,” an AI Platform that Predicts Clinical Outcomes from the Earliest Stages of Drug Discovery).
🟠 Enchant™ is a breakthrough multi-modal transformer model that breaks down the “data wall” 🧱 between preclinical and clinical research and development,
🟠Trained at scale across dozens of data sources and modalities, Enchant leverages abundant discovery-stage data with small amounts of human data to better predict key clinical properties of drug candidates, and
🟠 Enchant enables Iambic to eliminate years of preclinical R&D and improve clinical success.
Iambic has discovered so far two candidates to advance into the clinic:
👉 IAM-H1, a highly potent and irreversible tyrosine kinase inhibitor (TKI) that selectively targets HER2 and HER2 mutants, while sparing EGFR.
👉 IAM-C1, a potential first-in-class selective dual cell-cycle kinase CDK2/4 inhibitor to address unmet needs in terms of therapeutic window and treatment resistance in cell-cycle-driven cancers.
🐟 Arc Institute launches the virtual cell challenge
Virtual cells 🌐 are an emerging frontier at the intersection of AI and biology and a key goal of these cell state models is predicting cellular responses to perturbations. For this reason the Virtual Cell Challenge 🏎️ is being established to catalyze progress toward this goal (Virtual Cell Challenge: Toward a Turing test for the virtual cell). This recurring and open benchmark competition from the Arc Institute will provide an evaluation framework, purpose-built datasets, and a venue for accelerating model development (Arc Institute Launches Virtual Cell Challenge to Accelerate AI Model Development).
Regarding Arc’s atlas now, 10x Genomics and Ultima Genomics partnered with Arc Institute to accelerate the development of the Arc Virtual Cell Atlas (10x Genomics and Ultima Genomics partner with Arc Institute to accelerate development of the Arc Virtual Cell Atlas). More specifically, the Arc Institute launched in February 2025 the Arc Virtual Cell Atlas, a growing resource for computation-ready single-cell measurements, starting with data from over 3️⃣0️⃣0️⃣ million cells. The initial release of the Atlas was Arc’s first step toward assembling, curating, and generating large-scale cellular data to fuel new insights from AI-driven biological discovery.
The Atlas has two foundational datasets: the first is a 🆕, open source, perturbation dataset called Tahoe-100M, created by Vevo Therapeutics, comprising 1️⃣0️⃣0️⃣ million cells and mapping 6️⃣0️⃣,0️⃣0️⃣0️⃣ drug-cell interactions across 5️⃣0️⃣ cancer cell lines. The second dataset, scBaseCamp, is the first single-cell RNA sequencing dataset from public data to be curated and reprocessed at scale using AI agents. Arc mined observational data from more than 2️⃣0️⃣0️⃣ million cells representing 2️⃣1️⃣ different species sourced from public repositories, and processed them to a standardized form (Arc Virtual Cell Atlas launches, combining data from over 300 million cells).
The Arc Institute is also collaborating with Nvidia. In particular, a new AI model for biology, was recently released by Arc Institute and Nvidia, that can predict which mutations within a gene are likely to be harmful and even design small, eukaryotic genomes (▶️ Evo 2 Can Design Entire Genomes).
🐟 AstraZeneca’s boost in Bengaluru
Global biopharmaceutical leader AstraZeneca announced a ₹1️⃣6️⃣6️⃣ crore (a crore is used to symbolize 10 million) (One crore Indian Rupees, INR, equals roughly 133,514 US dollars as of April 27th, 2023) investment to expand its Global Hub in Bengaluru, reinforcing India’s growing role in global healthcare advancement (₹166 Crore AstraZeneca Boost in Bengaluru: 400 Jobs, Focus on AI & Health Tech). This strategic development adds 4️⃣0️⃣0️⃣ 🆕 positions, increasing the total workforce at the Bengaluru facility to nearly 1️⃣,3️⃣0️⃣0️⃣ professionals.
This is AstraZeneca’s second major expansion in India within a year, following the recent growth of its Global Innovation and Technology Centre (GITC) in Chennai. Together, the expansions bring the company’s total India workforce to close to 4️⃣,0️⃣0️⃣0️⃣ employees.
🐟 9Bio Therapeutics wants to revolutionize cancer treatment with its AI-Driven platform
9Bio Therapeutics a Québec biotech startup is developing a platform for designing drugs that target tumours and spare healthy tissues. The platform is an AI-guided structural biology platform that 9Bio plans to use to design therapies that selectively target tumours and spare healthy tissues (9Bio Therapeutics’ Philipe Gobeil wants to “resurrect” clinically validated cancer therapies that failed due to toxicity).
Their AI-guided structural biology platform integrates anatomical and biochemical contexts to design drugs that selectively target tumors. Their pipeline includes 2️⃣ discovery-stage therapies: 🔜 an antibody-drug conjugate (ADC) targeting chemoresistant tumor cells and 🔜 an ADC targeting tumor-subverted immune cells.
Furthermore, 9Bio Therapeutics was among the 10 Top 🔝 AI Drug Discovery Companies and Startups to Watch in 2025 by StartUs Insights. These 1️⃣0️⃣ drug discovery AI companies work on solutions ranging from computational discovery and bio-foundation models to therapeutic protein and molecule analysis, including 👉
LinkGevity: Longevity Science-led Drug Discovery
Examol, computational drug discovery solutions: AI-led Computational Drug Discovery
Helical AI: Bio-Foundation Models Platform
DenovAI Biotech: Therapeutic Protein Design
Orakl Oncology: AI-powered Techbio platform
AVAYL, AI for Medical Affairs: Medical Information-based Response Matching
Aevai Health: Intelligent and Personalized Chatbot for Biobanks
9Bio Therapeutics: AI-guided Structural Biology Platform
chAIron: AI-driven Molecules Analysis
Aureka Biotechnologies: Generative AI-based Therapeutic Design
🧃 Ainnocence launched BioSynthAI
On June 27, 2025, Ainnocence announced the launch of its BioSynthAI™ platform, an advanced AI solution for product discovery in the synthetic biology field, that leverages cutting-edge AI algorithms to optimize protein and enzyme activity through in-silico design, aiming to dramatically accelerate innovation across biopharmaceuticals and sustainable biotechnology (Ainnocence Launches BioSynthAI Platform for AI-Driven Synthetic Biology Product Discovery).
📒 SentinusAI is a de novo antibody and fusion protein engineering engine that performs antibody design and optimization based solely on sequences, delivering computational results in just a week and an optimized drug candidate within a month.
📒 The SentinusAI® platform is an AI-powered protein design engine that enhances antibody affinity and addresses challenging targets, including membrane and secreted proteins. It combines advanced epitope mapping with de novo hit generation, affinity maturation, off-target screening, humanization, and developability optimization.
📒 This comprehensive approach can virtually screen up to 10^10 antibody sequences within hours to days! The platform is also able to deliver a shortlist of candidates with a high wet lab hit rate, accelerating the development of life-saving therapies.
📒 SentinusAI’s platform designs various antibody formats, including full-length IgG antibodies, antibody fragments (Fab, scFv, VHH), bispecific and multispecific formats, and constructs for antibody-drug conjugates and CAR-T applications.
📒 It excels in designing bispecific antibodies targeting CD3, BCMA, and other cancer-related targets, as well as high-affinity antibodies for antibody-drug conjugates (e.g., Her2, CD22, CD30), addressing solid tumor penetration challenges and maximizing therapeutic index.
The company behind SentinusAI is Ainnocence (2021, US) that is a next-generation biotech company with a fast, self-evolving AI drug design platform consisting also of:
✳️ The CarbonAI is a de novo small-molecule and PROTAC design engine that holds the power to screen billions of compounds in mere days, optimizing multiple pharmacological properties simultaneously for lead generation and lead optimization.
✳️ And CellulaAI represents a cutting-edge AI system that leverages the power of AI to transform CAR-T therapy. By optimizing every aspect of the CAR-T design and production process, this engine aims to bring safer, more effective and personalized cancer treatments to patients worldwide (AINNOCENCE LAUNCHES CellulaAI™: A Pioneering AI Engine to Revolutionize CAR-T Therapy).
🧃 Causaly launched AI-powered competitive intelligence app 📲
Causaly just announced Pipeline Graph, giving pharma research and development teams a more effective and efficient way to do competitive intelligence earlier in drug discovery (Causaly Launches AI-powered Competitive Intelligence App for Early-Stage Drug Discovery).
🟠 Building on the company’s advanced AI platform and recent innovations such as Causaly Discover and Causaly Deep Research, Pipeline Graph unifies scientific research and market insights into a single, AI-native application designed specifically for pharmaceutical scientists, to make confident pipeline decisions faster before investing in costly drug development programs.
🟠 Pipeline Graph extends the capabilities of Causaly’s proprietary knowledge graph—which connects over 5️⃣0️⃣0️⃣ million facts and 1️⃣0️⃣0️⃣ million directional relationships, with more than 4️⃣ million new data points added monthly—by integrating competitive intelligence alongside biomedical data. This unified system:
Allows scientists to access internal and external structured and unstructured data in one platform.
Combines scientific literature, clinical trials, news, websites, and pipeline data to create a 360-degree view of the therapeutic landscape.
Seamlessly integrates into researchers’ existing workflows, eliminating the need for multiple disconnected tools and extensive manual searches.
In fact, pharmaceutical research teams face significant challenges in synthesizing fragmented competitive intelligence, since critical insights are often buried in 👉 emails, 👉 presentations, and 👉 static databases that lack real-time updates or comprehensive coverage—particularly for novel targets. Pipeline Graph resolves these pain points by offering a central, AI-native hub where researchers can rapidly connect the dots between early discovery, safety and efficacy data, and commercial strategy.
Causaly by using a hybrid approach that combines a best-in-class knowledge graph with the latest advances in generative AI, enables researchers to conduct deep, unbiased scientific exploration from target identification to biomarker discovery during drug development. Basically Causaly (London UK, 2017) offers a semantic AI-platform which reads collections of scientific articles and extracts causal associations through linguistic and statistical models.
In July 2023, the London startup Causaly raised 💲6️⃣0️⃣M, a Series B that will be going toward R&D and to continue building out its team (Causaly, an AI platform for drug discovery and biomedical research, raises $60M). Yiannis Kiachopoulos, the CEO who co-founded the company with CTO Artur Saudabayev, said that Causaly has already worked with 1️⃣2️⃣ of the world’s biggest pharmaceutical companies and some of the biggest names in medical research.
On March 31, 2025, Causaly announced 🆕 scientific AI agents that provide research teams with the industry's most comprehensive biomedical knowledge for drug discovery (Causaly Announces Agentic AI for Scientific Discovery). With agentic AI in Causaly Discover, life sciences teams can access, analyze, and synthesize information across the Causaly Knowledge Graph and internal and external data sources and researchers can tap into the broadest and deepest ecosystem of biomedical information to answer their research questions with unprecedented speed, accuracy, and transparency.
🧃 Ginko Bioworks stock rises
Ginkgo Bioworks Holdings (NYSE:DNA) stock climbed 🆙 3️⃣.8️⃣% following the announcement of a new collaboration with Hugging Face – The AI community building the future to open up high-quality biological datasets for ML applications (Ginkgo Bioworks stock rises on Hugging Face collaboration for AI drug development).
The partnership aims to make biological data more accessible to the AI community, potentially accelerating drug development, accordingly the companies released the complete 👉 GDPx functional genomics and 👉 GDPa antibody developability dataset series on their Hub. This release contains everything you need to explore molecular interactions in the cell—between genes, proteins, antibodies, and more—unlocking critical applications in biological research and drug discovery.
From transcriptomic response prediction to antibody property inference, these datasets support use cases like perturbation-response modeling, mechanism of action (MoA) characterization, and make it possible to build models of perturbation responses and antibody developability (Accelerating AI for Drug Discovery: Ginkgo’s GDPx Functional Genomics and GDPa Antibody Developability Dataset Series).
🧃 OutSee raised 💷 1️⃣.8️⃣M
The Cambridge-based predictive genomics drug discovery company OutSee has raised £1️⃣.8️⃣M in seed funding 🌱. The company is utilizing AI to uncover therapeutic targets directly from genomic data, and plans to expand its in-house target pipeline and partner with major pharma and biotech firms (OutSee raises £1.8M to unlock predictive genomics drug discovery with AI platform Nomaly).
OutSee is developing and applying innovative computational approaches to genomics for drug target discovery and precision medicine. Nomaly is their platform that predicts disease and phenotype directly from a single genome ab initio, not by pattern matching known genetic associations, but from fundamental molecular and biology. The model it’s not trained to look for familiar variants or correlations, instead it uncovers the underlying biological mechanisms that actually drive and modulate disease.
💟 Mayo Clinic's StateViewer now identifies 9️⃣ types of dementia, including Alzheimer's, using a single FDG-PET scan with an impressive 8️⃣8️⃣% accuracy
Mayo Clinic's AI marvel, StateViewer, now identifies 9️⃣ types of dementia, including Alzheimer's and related diseases, using a single FDG-PET scan with an impressive 8️⃣8️⃣% accuracy. In fact, StateViewer accelerates diagnosis by analyzing brain glucose usage, making it a game-changer for clinics without neurology specialists (Mayo Clinic’s StateViewer AI Revolutionizes Dementia Detection with 88% Accuracy in a Single Scan).
More specifically, StateViewer applies a neighbor matching algorithm to detect the presence of 9️⃣ different neurodegenerative phenotypes (An FDG-PET–Based Machine Learning Framework to Support Neurologic Decision-Making in Alzheimer Disease and Related Disorders). The ML performance of this framework was evaluated in a discovery cohort by nested cross-validation and externally validated in the Alzheimer's Disease Neuroimaging Initiative.
👉 The discovery cohort contained 3,671 individuals with a mean age of 68 years and consisted of 49% reported female. The model framework was able to detect the presence of 9️⃣ different neurodegenerative phenotypes with a sensitivity of 0.89 ± 0.03 and an area under the receiver operating characteristic curve of 0.93 ± 0.02.
💟 MindRank announced positive phase 2b results
MindRank, a clinical stage AI-empowered drug discovery company, announced positive topline results from a phase 2b clinical trial of its proprietary AI-designed oral GLP-1 receptor agonist (GLP-1RA), MDR-001, in adults with obesity or overweight in China (MindRank Announces Positive Phase 2b Results).
In this 24-week, randomized, placebo-controlled study, MDR-001 demonstrated clinically meaningful, dose-dependent weight reduction. Participants receiving MDR-001 achieved mean body weight reductions ⬇️ ranging from 8️⃣. 2️⃣ % to the 1️⃣0️⃣. 3️⃣% (7.4-9.2 kg) compared to 2️⃣. 5️⃣% (2.4 kg) in the placebo group (p<0.00001).
In addition to weight reduction, MDR-001 delivered significant improvements in key cardiometabolic 🤸 markers, including waist circumference, blood pressure, and lipid profiles, underscoring its potential as a comprehensive metabolic therapy.
Overall MDR-001 was well tolerated, with no treatment-related serious adverse events (SAEs) ☠️ reported. The most common treatment-emergent adverse events (TEAEs) were mild to moderate gastrointestinal symptoms, such as nausea 🤢, vomiting 🤮, and diarrhea 💩, which were predominantly observed during the initial 6-week dose-escalation period and resolved within 1 to 5 days.
Importantly, hepatic safety analyses showed no evidence of transaminase elevation, even among approximately 2️⃣0️⃣% of participants with pre-existing liver impairment. In fact, ALT and AST levels were significantly reduced ⬇️ in the MDR-001 treatment groups compared to placebo. Additionally, no clinically relevant increases in heart 🫀 rate were observed.
The overall discontinuation rate due to TEAEs was only 0.8%.
MindRank (2021) is a 🔝 AI-driven biotech company in China with focus on “undruggable” targets, with R&D centers in China and UK, that in 2021 was Invited to Join the “NVIDIA Startup Acceleration Program”. Its AI platform consists of 👉 Molecule Pro (A molecule design and generation platform), 👉 Molecule Dance (A molecular dynamics platform to simulate protein movements) and 👉 PharmKG (A biomedical knowledge graph to assist drug discovery).
Apart from the MDR-001 candidate for Obesity and Type 2 Diabetes, currently they are also working on the following assets:
🔷 MRANK-106, Undisclosed, Oncology (IND-enabling),
🔷 In November 2023, they started conducting IND-enabling studies for their “first-in-class” preclinical candidate MRANK-106.
On March 07, 2025, FDA cleared the company's Investigational New Drug (IND) application for MRANK-106, a first-in-class, orally available dual inhibitor of WEE1 and YES1 kinases, for the treatment of pancreatic cancer, small cell lung cancer, ovarian cancer, breast cancer, and colorectal cancer (Dual WEE1/YES1 Kinase Inhibitor MRANK-106 Secures FDA IND).
🔷 MRANK-108, Undisclosed, Oncology (IND-enabling),
In November 2023, they initiated the IND-enabling studies for their preclinical candidate MRANK-108.
🔷 MRANK-103, Molecular Glue, Oncology
On July 29, 2024, MindRank announced that its molecular glue project, MRANK-103, targeting the MYC signaling pathway and designed to regulate eukaryotic translation and transcription factors, has completed the nomination of preclinical candidate and begun the IND-Enabling studies (MindRank Begins IND-Enabling Studies of its AI-Designed Molecular Glue Drug Pipeline MRANK-103).
🔷 MRANK-023, Molecular Glue, Undisclosed (Optimization),
🔷 MRANK-111, Undisclosed, Obesity (Optimization) and
🔷 MRANK-603, Undisclosed, Undisclosed (Optimization).
Moreover, they have 10 more programs (Undisclosed) in co-development and they have designed as a service for public biotech two assets:
EMRANK-016, Allosteric Inhibitor, Oncology (IND-enabling) and
EMRANK-189, Undisclosed, Undisclosed (Optimization).
On January 19, 2024, MindRank and Henlius (2696.HK) announced a strategic collaboration to jointly develop two innovative healthcare solutions. One is an AI-driven antibody-drug conjugates (ADCs) drug discovery platform, which will expedite the research and development of next-generation ADCs with better clinical efficacy. The other solution is an AI-driven anti-aging therapy platform, which will be developed by leveraging MindRank's biomedical AI large language model (MindRank and Henlius (2696.HK) Announce Strategic Collaboration for AI-Driven Development of Next-Gen Antibody Drug Conjugates (ADCs) and Anti-Aging Therapy Platform).
On April 2, 2024, MindRank announced a strategic partnership with BioMetas Co, Ltd, based in Shanghai, for preclinical drug research and development to advance the construction of a next-generation AI+ drug research and development platform (MindRank and BioMetas Announce Strategic Partnership to Establish One-Stop Service Platform for AI+ Drug Research and Development).
Finally, on May 21, 2024, MindRank and Zhejiang Hisun Pharmaceutical Co, Ltd (“HISUN”, 600267.SH) signed a strategic cooperation agreement on AI-empowered small molecule drug discovery (MindRank and HISUN (600267.SH) Forge Strategic Partnership to Advance AI-Empowered Innovative Drug Discovery).
💟 Biomedical AI-Q Research Agent by NVIDIA
NVIDIA has developed the Biomedical AI-Q Research Agent to assist drug development scientists to rapidly review available literature, draw complex hypotheses, and then hand off the uncovered protein targets to a virtual screening agent. When performed manually, this process traditionally would be time consuming and cumbersome, involving days of reading and summarizing papers (Advancing Literature Review & Target Discovery With NVIDIA Biomedical AI-Q Research Agent Blueprint).
This Biomedical AI-Q Research Agent Developer Blueprint incorporates elements from the RAG Blueprint as well as the newly released NVIDIA AI-Q NVIDIA Blueprint. Additionally, NVIDIA’s approach leverages aspects of the BioNeMo Virtual Screening Blueprint to take the hypotheses built by the reasoning agent and utilize novel small molecule candidates for a specific protein target.
💟 Paige announced AI integration with Roche’s navify
Paige announced AI integration with Roche’s platform navify® Digital Pathology (Diagnostic tools will be embedded into Roche's navify system to support digital pathology workflows).
This integration marks another important step toward making AI more accessible in routine clinical workflows—empowering pathologists with powerful tools to support diagnostic efficiency and more informed decisions.
💟 Google Deep Mind released AlphaGenone
Google DeepMind unveiled another major leap forward with 👉AlphaGenome, that is a DL model designed to predict how genetic variants, especially those tricky non-coding ones (which make up over 9️⃣8️⃣% of genetic variation), influence genome function.
🟠 Trained on vast datasets of human and mouse genomes, AlphaGenome tackles the challenging task of interpreting variant impacts at a scale and resolution never achieved before.
🟠 It simultaneously predicts thousands of genomic features across diverse modalities, including gene expression, splicing, chromatin accessibility, histone modifications, transcription factor binding, and even complex chromatin contacts.
🟠 Impressively, it handles massive genomic contexts (1 million DNA base pairs!) at single-base resolution.
🟠 It predicts genomic activity across multiple modalities with unprecedented accuracy, outperforming existing models in 24 out of 26 evaluations.
🟠 Unlike specialized models that focus narrowly, AlphaGenome offers a unified view, predicting variant effects across diverse biological mechanisms.
🟠 Finally, it's already demonstrating the ability to elucidate complex, clinically relevant genomic mechanisms (such as those near the TAL1 oncogene), opening doors to enhanced variant interpretation and therapeutic targeting.