TechBio: Latest Trends 😺
TechBio: Latest Trends 😺
MetaphysicalCells: A newsletter about Science, Technology and AI Drug Discovery
“The computer scientist Donald Knuth was struck that
«AI has by now succeeded in doing essentially everything that requires ‘thinking»
but has failed to do most of what people and animals do ‘without thinking’—that, somehow, is much harder!”
By Nick Bostrom, Superintelligence: Paths, Dangers, Strategies
News ✴️
🏖️ Next week Thibault GEOUI from Charles River Laboratories will give a presentation at "world Class Corporate Analytics & Artificial Intelligence 2024" organized by World Class Business Leaders. Join him at link.
🏖️ LabGenius just announced that their paper on “Best practices for machine learning in antibody discovery and development” is published in Drug Discovery Today. Please have a look if you are interested in understanding:
- How to generate consistent, high-quality data for antibody drug discovery projects,
- Process the data in a way to make it suitable for machine learning, and
- Evaluate machine learning methods appropriately to ensure good performance in drug discovery programs.
🏖️ NVIDIA features QSimulate’s simulation technology! QUELO-G is a revolutionary lead optimization solution delivering quantum accuracy for small molecule drug discovery. Using the latest GPU hardware, simulations can be performed in a fraction of a day.
🏖️ Transomics + AI is the recipe needed to uncover new therapeutic possibilities.
Pepper Bio’s CEO Jon (Yu) Hu gives a deep dive into the process that led to the discovery of two inhibitors (and a $135M in-licensing deal) in Drug Target Review: Transomics’ impact on a promising liver cancer therapeutic.
🏖️ An AI leap into chemical synthesis: EPFL scientists introduce ChemCrow, a large language model-based AI system that revolutionizes chemistry by integrating 18 advanced tools for tasks like organic synthesis and drug discovery. ChemCrow streamlines complex processes in chemical research, making it more efficient for experts and novices alike.
🏖️ Nabla Bio Secures $26M Series A Financing and Collaborations with AstraZeneca, Bristol Myers Squibb and Takeda for Generative Protein Design.
🏖️ AI is a viable alternative to high throughput screening: a 318-target study.
🏖️ Oregon Therapeutics & Lantern Pharma Launch Strategic AI Collaboration to Optimize Development of First-In-Class Drug Candidate XCE853 – A Potent Inhibitor of Cancer Metabolism.
Menten AI 🧙♀️
Menten AI, Inc. in San Francisco is a biotechnology company with a generative AI platform for design and optimization of peptide macrocycles. They developed a platform for protein design powered by ML and quantum computing, by utilizing proprietary algorithms of quantum optimization which they believe can significantly improve the accuracy of drug discovery while reducing cost and development time. The Menten AI platform designs peptide macrocycles in a larger chemical space (1050 compared to 109 for high throughput platforms) and in record time (8 weeks vs 1-2 years in industry). This equates to 10,000x greater sampling power in ⅙ the time with >50% in vitro hit success compared to ~9% in industry.
Menten AI is using the NVIDIA cuQuantum software development kit to accelerate their quantum circuit simulations on GPUs and support its quantum work. Co-founded by Hans Melo and Vikram Mulligan, Menten AI has partnerships with D-Wave supercomputer, IBM-Q and a collaboration with Xanadu.
On May 28, 2024, Menten AI announced the completion of a previously undisclosed research collaboration and licensing agreement with Bristol Myers Squibb applying Menten AI’s platform for optimization of cyclic peptides. Under the research collaboration, Menten AI and Bristol Myers Squibb leveraged Menten AI’s generative AI platform and team’s expertise to optimize the biochemical properties of certain peptide macrocycles. Together, the teams were able to explore a more expansive chemical space and identified new amino acid modifications to improve the desired properties.
Backed by top-tier VCs including Uncork Capital, Khosla Ventures, Social Impact Capital and Y-Combinator, Menten AI has raised a total of $4M and is funded by 8 investors. Sahin Boydas and big_bets are the most recent investors.
Take a look at the most recent publications by their team members:
How to Design Peptides (Chemokine-Glycosaminoglycan Interactions: Methods and Protocols),
Stabilizing proteins, simplified: A Rosetta-based webtool for predicting favorable mutations and
Big_Bets 🧙
Big_Bets is a venture capital and private equity company in Brazil 🇧🇷 that invests exclusively in software-intensive companies, and the partners are Alexandre Mello and Luiz Guilherme Manzano. The investments of Big_Bets are highly agnostic, covering areas like financial services (Conta Simples, Ume and Justos), identity (Unico), pharmaceutical industry (Menten AI and Pepper Bio), diagnostics (Huna), media (Festalab, Prota, and Newtail), primary education (Studdy), energy distribution (Lemon), devtools (Lazy AI, DataMilk, NuMind, Croct), among others. Regarding TechBio, and apart Menten AI, Big_Bets has invested so far in:
Pepper Bio (Boston US, 2018) is a TechBio company that wants to reduce costs by getting drug developers to the right targets faster, using what it calls a “transomics” approach, which maps data across all the -omics layers and translates information in a way similar to Google maps that integrates various layers of data. Accordingly, Pepper’s COMPASS platform decodes layers of biological maps—genomics, transcriptomics, proteomics and phosphoproteomics—providing the fastest route for drug discovery. Just a month ago, they announced an oversubscribed $6.5M seed round led by NFX with participation from Silverton Partners, Merck Digital Sciences Studio, Mana Ventures, Tensility Ventures, VSC Ventures, and more.
Huna AI develops innovative solutions to support the early diagnosis of cancer through routine, low-cost tests. With faster diagnoses, in the early stages, Huna helps to increase the chances of a cure, optimizes healthcare system resources and reduces treatment complications, generating an improvement in patients’ quality of life. They analyze routine exams using their exclusive technology so managers and healthcare professionals can have access to a risk classification for each patient, contributing to better clinical decisions and early diagnoses.
In 2023, Daniella Castro, CTO & Co-founder at Huna, was one of the confirmed speakers at Brazil at Silicon Valley – an annual conference that connects brazilian and american entrepreneurs and leaders in the heart of Silicon Valley.
In 2024, Huna was a finalist for the Aurora Tech Award, a worldwide prize for women founders of tech startups.
News ✨
🍍 DeepCure's Automated Synthesis Transforms AI Drug Designs into Testable Compounds. The company's Inspired Chemistry platform performs automated, multi-step synthesis of small molecules including those designed using AI.
🍇 AIGENDRUG, is an AI drug discovery research company that provides a new way of exploring the chemical, genetic and disease space with cutting edge AI technologies.
🍉 BCG Announces GenAI Collaboration with Merck
🍓 The royalsociety published "Science in the Age of AI - How AI is changing the nature and method of scientific research," and it's a must-read for everyone interested in AI & science.
🍑 Generative AI in Healthcare System and its Uses. Complete Guide
Weave Bio 🧝
Weave Bio (US) is modernizing your IND process (Investigational New Drug Application). Conventional IND submissions can take months for your team to draft but with Weave’s AI-native platform you can draft your IND in a day, leveraging Weave’s in-platform process for quality review and significantly enhancing editing and refinement capabilities.
On May 30, 2024, Weave Bio announced the public release of its flagship offering, AutoIND, and $10M in seed funding co-led by Innovation Endeavors and Magnetic Ventures. AutoIND, part of Weave’s AI-native platform, allows drug developers to build a dynamic, living record of their therapeutic candidates while removing much of the manual work associated with summarizing data from studies.
AutoIND automatically generates content based on your own data, as well as regulatory guidelines and public information. In this way, the solution offers a robust process for quality review, which builds a living record of the development process for a therapeutic candidate while also significantly enhancing editing capabilities and oversight for teams. More importantly, while conventional IND preparation processes can take more than a year, with AutoIND you can produce a first draft of an IND in a single day and reduce the overall time to generate a final IND submission by over 50%!
So far, the company doubled the development pace of AutoIND and established a roadmap for its broader AI-native Regulatory Automation and Lifecycle Management (ReALM) platform.
Biolojic 🧝♂️
Biolojic Design (Israel, US) is utilizing ML to create superior drugs by programming antibodies to become dynamic functional switches that affect biology in new ways. Biolojic’s AI-designed antibodies provide a precise molecular effect, in the desired micro-environment, at the right time, leading to greater efficacy and lower toxicity, while maintaining impeccable developability, half-life, stability and safety of human antibodies.
In particular, their platform mimics the way the human immune system makes antibodies, allowing them to create personalized smart drugs that act like nanobots in the body. Unlike conventional drugs, which behave the same way everywhere, these programmable antibodies can sense the micro-environment in a specific patient at a specific time and respond dynamically.
For example, AU-007 is the first-ever computationally designed antibody to enter human clinical trials and was designed by Biolojic to precisely bind to pre-defined epitope on interleukin-2 (IL-2) to redirect IL-2 toward cells that express dimeric receptors (T effs, NKs, NKTs), and away from cells that express the trimeric receptor (T regs, vascular endothelium). Thus, AU-007 unlocks the full potential of IL-2 to fight cancer: it starves the cells that protect the tumor and nourishes those cells that kill the tumor.
On April 24, 2024, Biolojic Design announced that Nektar exercised its license option to develop a program resulting from the companies’ research collaboration initiated in 2021. The research collaboration was established to design and test agonistic antibodies targeting tumor necrosis factor receptor type II (TNFR2), a novel, biologically-validated disease target implicated in a wide range of autoimmune diseases that has been difficult to drug with conventional medicines.
On June 3, 2024, Biolojic Design Entered into a Multi-Target Drug Discovery Collaboration with Merck KGaA, Darmstadt, Germany. Merck will leverage Biolojic Design’s AI capabilities to create potential best-in-class multi-specific antibodies that uniquely modulate biology and Biolojic Design will receive a low double-digit million euro upfront payment and is eligible for up to €346M in discovery, development, regulatory and commercial milestones, in addition to royalty payments on net sales.
Biolojic Design has raised about $30M to date.
BioLib 🧙♂️
BioLib (Denmark, 2019) is the library of biological data science where you can access bioinformatics tools securely without leaving the browser. BioLib applications run on your local machine or using secure cloud computing. Once you generate insights from your data using bioinformatics tools on BioLib, you can run tools through the graphical user interface with just a few clicks or, if you are comfortable with programming, you can call them directly from existing notebooks or data pipelines.
You do not need to be a front-end developer or DevOps-wizard to make your tools accessible to a wide audience. You can simply upload your source code, define the user inputs using a graphical interface or the YAML configuration file, and the application is ready to go! By integrating AI with the world of biological data science they harness the power of ML and innovative technologies to revolutionize scientific discovery and drug development.