Weekly Tech+Bio Highlights #9
ALSO: A New Foundation Model Learns Holistic Biology; Accidental Breakthrough in Aging Research; The Promise of 3D In Vitro Models for AI in Drug Discovery; 10 Notable Biotech VC Rounds
Hi! I am Andrii Buvailo, and this is my weekly newsletter, ‘Where Tech Meets Bio,’ where I talk about technologies, breakthroughs, and great companies moving the biopharma industry forward.
If you've received it, then you either subscribed or someone forwarded it to you. If the latter is the case, subscribe by pressing this button:
This Week’s Sponsor: Intelligencia AI
Upcoming Webinar: Elevate and De-Risk Your Drug Portfolio Strategy With AI | Aug. 28 @ 11 AM EDT
Risks, critical decisions, and subjectivity surface at various points in drug development. So, how do data-driven insights and AI reduce risk and improve quantitative rigor? This conversation will cover perspectives and approaches to elevate portfolio strategy capabilities and de-risk drug development. You'll hear industry leaders who have used decision sciences and AI to empower operational and business goals. The panelists will share how they have used AI and insights tools to validate and calibrate the probability of technical and regulatory success (PTRS), support asset prioritization, and inform pipeline performance and strategic decisions about asset development. In addition, learn how real-time updates on your portfolio and competitors can level up your acquisition and licensing efforts.
Register for the webinar today.
You’ll hear from:
John McGrath - Project Management Professor/Course Director MSc Project Management
Panos Karelis - Director of Customer Experience and Insights | Intelligencia AI
Ivan Kugener - VP, Head of Global Portfolio Management, Valuation and Analytics | Merck KGaA
Joshua Hattem - Principal | ZS
Now, let’s get to this week’s topics!
News Highlights
🧠 US-based Synchron integrates generative AI into its brain-computer interface (BCI) system, enabling non-verbal individuals with conditions like ALS to generate text and audio using their thoughts, enhancing communication speed and efficiency.
🔬 Evaxion Biotech showcased advancements in its AI-Immunology™ platform at ISMB 2024, highlighting enhanced peptide-MHC prediction accuracy through new deep-learning frameworks and strategies, aiding vaccine target discovery for cancer and infectious diseases.
🔬 Scientists create first mouse model with complete, functional human immunesystem and a human-like gutmicrobiome that is capable of mounting specific antibody responses.
💰 Israeli startup CytoReason raised $80 million from investors including NVIDIA and Pfizer to scale its AI disease models, enhancing its molecular and clinical database, and expanding into new disease areas and markets.
🧬 Accidental breakthrough in aging research from Duke–NUS Medical School reveals that blocking IL-11, an inflammation-promoting protein, extends the lifespan of mice by 25%, improves metabolism, and reduces frailty, suggesting potential for anti-aging therapies in humans (Nature, July 17).
🔬 310 AI launched MPM4, a generative text-to-protein sequence model, advancing the "Programming Problem" by designing novel proteins from specified functions. MPM4 integrates with tools for validation through structure modeling and function prediction, and released a repository of 1000+ high-quality AI-generated protein sequences.
🔬 Exscientia acquired full rights to the CDK7 inhibitor GTAEXS617 from GT Apeiron, ahead of Phase 1 dose escalation data readout in late 2024. The ELUCIDATE trial will assess '617 as a monotherapy and in combination with standard therapies, starting with HR+/HER2- breast cancer.
💰 Sotio agreed to pay Biocytogen up to $325.5 million in upfront and milestone fees to create bispecific antibody-drug conjugates (ADCs) using Biocytogen's transgenic mice-generated bispecific antibodies, enhancing Sotio's ADC capabilities for cancer treatment.
💰 Barcelona-based Integra Therapeutics secures €10.5 million from the European Commission through the EIC Accelerator program to advance its FiCAT gene writing platform, supporting development of therapies for rare diseases, autoimmune disorders, and oncology.
🚚 SkyCell AG partnered with Validaide to enhance pharmaceutical supply chain decision-making by integrating SkyCell's SkyMind software with Validaide's lane risk assessment and CO₂e comparisons, aiming to reduce the $35 billion annual losses from temperature-controlled logistics failures, while cutting costs, risks, and environmental impact.
🔬 Brainomix and Boehringer Ingelheim collaborate to enhance care for fibrosing lung disease in the US using Brainomix’s FDA-cleared e-Lung imaging software, aiming to shorten the diagnosis time from up to 2 years and expedite treatment initiation.
A New Foundation Model Learns Holistic Biology
While AI models like AlphaFold can predict biomolecular structures, drug failures often occur due to incorrect targets or patient selection, not structural issues.
San Francisco-based NOETIK has released an insightful technical report on the Oncology Counterfactual Therapeutics Oracle (OCTO), an AI model that uses a transformer-based architecture trained on multimodal data from patient tumors, including protein staining, gene expression, DNA sequencing, and structural markers.
Its innovative design integrates diverse biological data to predict patient-specific treatment responses for complex diseases like cancer.
With features like structured multimodal masking and visual prompting, OCTO can infer intricate biological relationships and simulate therapeutic interventions, aiming to enhance precision medicine and streamline drug development.
Read a detailed blog post by NOETIK explaining how OCTO works, but if you need a brief reference, here is a very short technical breakdown:
Transformer-Based Architecture:
Utilizes transformer layers to process and integrate large-scale multimodal data.
Capable of handling complex sequences of biological information from various sources.
Multimodal Data Integration:
Trained on diverse data types: multiplex protein staining, spatial gene expression, DNA sequencing, and structural markers (e.g., H&E staining).
Data is encoded into a unified representation or embedding space, allowing cross-modality analysis.
Structured Multimodal Masking:
Employs a novel form of masked data modeling where nearly all tokens in each modality are masked out.
Forces the model to learn relationships across different data types by revealing minimal information.
Visual Prompting:
Model receives partially revealed images as input (visual prompts) to infer the remaining biological data.
This technique helps in predicting the effects of biological perturbations (e.g., gene knockouts).
Tokenization of Data:
Converts multiplex fluorescence images into discrete color channels, then into "patchified" tokens.
Spatial gene expression data is tokenized by gene and cell location, creating a detailed spatial map.
Pre-Training:
Pre-trained on a dataset containing 20 billion tokens across 128 GPUs.
Capable of including up to 16,000 tokens in its context window during inference.
High-Dimensional Output:
Produces 16-channel images capturing protein expressions in tumor sections.
Outputs are high-resolution and therapeutically relevant, enabling detailed biological simulations.
Counterfactual Simulations:
Conducts in silico experiments by manipulating input data to simulate hypothetical treatment scenarios.
Predicts the biological impact of changes, such as altering gene expression or protein levels.
Biological Knowledge Extraction:
Uses heavy masking during training to ensure the model learns to infer strong relationships from sparse data.
Capable of reconstructing high-resolution images and predicting protein co-expression and cell types from minimal inputs.
Self-Supervised Learning:
Trains on unlabeled data using structured masking to learn robust representations.
Enhances the model's ability to generalize and infer biologically relevant features without explicit labels.
The Promise of 3D In Vitro Models for AI in Drug Discovery
In an interview with Dr. Jan Lichtenberg, Co-Founder and CEO of InSphero, a renowned biotech enterprise specializing in 3D cell culture technologies, for the Medicine Maker, the discussion centered around the pivotal role of scalable 3D in vitro models in AI-driven drug discovery.
Dr. Lichtenberg highlighted how these advanced models are enhancing the predictive accuracy of preclinical studies and providing the extensive datasets required for machine learning (ML) and AI algorithms.
Key Points
Scalable 3D in vitro models can be produced in large quantities, providing consistent and extensive datasets necessary for effective AI and ML training. These models leverage primary cells without the inconsistencies of donor changes due to the miniaturization of 3D spheroids. The uniformity in size, shape, and cell composition of these models ensures accurate representation of tissue responses and compound effects, enhancing the quality of data for ML.
The high-quality microplates used for delivering 3D tissues are compatible with standard liquid-handling robotics, minimizing the risk of loss during pipetting steps. These plates also support advanced imaging capabilities, enabling high-content imaging and cell painting to generate deep datasets. Additionally, the integration of biochemical assays and various omics readouts further enriches the data, providing a comprehensive dataset for AI training.
The scalability of 3D in vitro models allows for studies across diverse donor populations, adding an important dimension of patient diversity to the data. This broad range of biologically relevant readouts, combining functional biochemical assays, imaging assays, cell painting, and omics techniques, generates multidimensional data that is crucial for training AI systems effectively.
Accidental Breakthrough in Aging Research: Blocking IL-11 to Extend Lifespan
Exciting news in the field of aging research! A recent study published in Nature has unveiled that blocking a protein known as IL-11 can significantly extend the lifespan of middle-aged mice by about 25%. This discovery opens up potential avenues for human longevity treatments.
Key Highlights:
🔹 IL-11 and Inflammation: IL-11 is an immune protein that promotes inflammation. Chronic inflammation is a key contributor to many age-related diseases. By blocking IL-11, researchers observed improved metabolism, reduced frailty, and increased lifespan in mice.
🔹 Human Implications: IL-11 exists in humans, and drug candidates targeting this protein are already undergoing trials for cancer and fibrosis. This positions IL-11 inhibitors as promising candidates for future longevity treatments.
🔹 Accidental Discovery: The connection between IL-11 and aging was discovered when higher levels of IL-11 were found in older rats during routine assays. Further tests confirmed that older mice consistently had higher IL-11 levels across various tissues.
🔹 Experimental Results: In the study, genetically modified mice lacking IL-11 lived healthier and longer lives. Additionally, using an antibody to block IL-11 in older mice led to similar health benefits.
Technical Aspects:
🔸 Healthspan vs. Lifespan: Blocking IL-11 not only extended the lifespan of mice but also improved their healthspan – the period during which they remain healthy. This contrasts with some treatments like rapamycin, which, while extending lifespan, can have adverse effects on healthspan.
🔸 Potential Human Trials: While the results in mice are promising, translating these findings to humans requires further research. Clinical trials focusing on specific age-related conditions, such as muscle mass loss, could provide quicker and more interpretable results.
Next Steps:
🧬 Validation: The next phase involves testing IL-11 inhibitors in mice with diverse genetic backgrounds and across multiple laboratories to ensure reproducibility.
🧬 Focused Clinical Trials: Instead of long-term lifespan studies, shorter trials targeting specific age-related conditions may expedite the development of IL-11-based therapies.
The journey towards effective longevity treatments is complex, but discoveries like this bring us closer to understanding the intricate biology of aging and developing practical solutions to enhance human health and lifespan.
10 Notable Biotech Companies With Recent Major VC Rounds (+bonus insight)
As the biotech sector recovers to pre-COVID investment levels, venture capital firms have become more selective in their funding choices. In 2024, private U.S. drug companies raised $3.8 billion in the first quarter, aligning with the $15 billion annual investment seen last year. This marks a return to the investment levels of 2018 and 2019, after a spike during the pandemic in 2021 and 2022. Venture capital firms are also actively raising new funds, with several new and established firms launching significant funding rounds.
However, there is a notable shift in investment focus.
Preclinical companies are receiving less funding, while larger private rounds are becoming more common. Clinical-stage companies, particularly those with experienced management teams, are more likely to succeed in pricing IPOs. The broader economic environment and the Federal Reserve's interest rate policies are influencing these investment patterns. As a result, investors are increasingly considering mergers and acquisitions (M&A) and partnerships as viable exit strategies, given the fluctuating IPO market.
Read also: 13 Publicly Traded Biotechs Utilizing AI-based Research Platforms (+ 2 Upcoming Public Debuts)
Below we list some of the notable venture capital rounds lately, with a focus on innovative technological platforms or breakthrough biology:
Asceneuron SA
The company secured $100 million in Series C funding to develop tau-targeting therapies for neurodegenerative diseases, with a focus on phase 2 clinical trials for Alzheimer’s disease. The company’s lead candidate aims to reduce the accumulation of tau protein tangles in the brain, which are associated with the progression of Alzheimer’s. Asceneuron’s therapies are designed to modify disease progression and improve cognitive function.
AusperBio
AusperBio raised $37 million in Series A funding to advance its targeted oligonucleotide therapies, particularly for hepatitis B. The company’s innovative platform designs and develops oligonucleotide-based treatments that target viral RNA, aiming to achieve functional cures for chronic infections. AusperBio’s therapies offer a new approach to treating viral diseases by directly targeting the genetic material of the virus.
Cardurion Pharmaceuticals
The company raised $260 million in a Series B funding round to develop therapies for cardiovascular diseases. Their lead program is in phase 2 clinical trials for treating heart failure. Cardurion’s approach involves innovative therapies that target the underlying mechanisms of heart disease to improve cardiac function and patient health.
CatalYm GmbH
CatalYm received $150 million in Series D funding to advance its immunotherapy pipeline, focusing on a phase 2b clinical trial for solid tumors. The company’s lead candidate is a novel immune-modulating antibody designed to enhance the body’s immune response against cancer. CatalYm aims to improve the efficacy of existing cancer immunotherapies by overcoming immune suppression within the tumor microenvironment.
Draupnir Bio
This biotech secured €12 million ($13.1 million) in seed funding to develop targeted protein degradation therapies for heart disease. The company’s preclinical research focuses on using small molecules to selectively degrade disease-causing proteins, potentially offering a new approach to treating cardiovascular conditions. Draupnir Bio’s technology aims to address the root causes of heart disease at the molecular level.
Element Biosciences
The prominent deal included US-based Element Biosciences, which raised $277 million in a Series D funding round to further develop and commercialize its innovative DNA sequencing and multi-omics technologies.
The company’s flagship product, AVITI™, is a benchtop DNA sequencer that offers high accuracy and affordability, positioning it as a competitor to Illumina. Element Biosciences plans to launch AVITI24™, a next-generation system that integrates sequencing with cyto-profiling, enabling simultaneous analysis of DNA, RNA, proteins, and cell structure within single cells, thus providing comprehensive biological insights.
HepaRegeniX GmbH
This German biotech received €15 million ($16.4 million) in Series C funding to further its regenerative therapies for liver diseases. The company’s lead candidate is in phase 2a clinical trials for liver regeneration. HepaRegeniX’s approach involves activating specific pathways that promote liver cell regeneration and repair, offering a potential treatment for conditions like liver cirrhosis.
IDEAYA Biosciences
The company secured $263 million through a public offering to advance its precision medicine programs, with a particular focus on a phase 2 clinical trial for ocular melanoma. The company specializes in synthetic lethality, a targeted therapy approach that exploits cancer-specific genetic vulnerabilities. IDEAYA’s precision medicine platform integrates molecular diagnostics with targeted therapies to improve patient outcomes.
Pan Cancer T
Pan Cancer T raised €4.25 million ($4.65 million) in seed extension funding to develop its TCR (T-cell receptor) therapies for solid tumors. The company’s preclinical programs focus on engineering T-cells to recognize and attack cancer cells more effectively. Pan Cancer T’s platform aims to provide personalized and potent immunotherapies for cancer patients.
Scorpion Therapeutics
The company raised $150 million in Series C funding to support its precision oncology programs, including phase 1/2 clinical trials for solid tumors. The company uses a proprietary drug discovery platform to develop targeted therapies that address specific genetic mutations in cancer. Scorpion’s approach aims to deliver highly selective and potent treatments with fewer side effects.
For a broader picture of the recent funding rounds and tech+bio trends in the first half of 2024, read one of the previous newsletters: