TechBio: Latest Trends
TechBio: Latest Trends
AI/ML tools, startups, companies and investors
Welcome back 🌻 to another edition of MetaphysicalCells on AI drug discovery news.
Today’s Quote
“We enjoy rest and relaxation, but our bodies are still those of endurance athletes evolved to walk many miles a day and often run, as well as dig, climb and carry. We love many comforts, but we are not well adapted to spend our days indoors in chairs, wearing supportive shoes, staring at books or screens for hours on end.”
By Daniel E. Lieberman, The Story of the Human Body: Evolution, Health, and Disease
News 📬
📩 Join Thibault Géoui, PhD (Head of Science Analytics at Charles River Laboratories) on April 8th at 2 PM for a thought-provoking discussion on AI in Healthcare, to explore both the opportunities and challenges AI poses within the healthcare sector. The conversation will offer an holistic perspective on this topic, providing a critical analysis for all interested in the field. Register ➡️ AI limitations in healthcare, by health-healthcare-hackers.
GSK and AI Drug Discovery
GSK plc is a British multinational pharmaceutical and biotechnology company with global headquarters in London. It was established in 2000 by a merger of Glaxo Wellcome and SmithKline Beecham, which was itself a merger of a number of pharmaceutical companies around the Smith, Kline & French firm.
In 2023, GSK invested heavily in harnessing data and AI for early-stage drug discovery, with a goal of hiring about 100 people to work on the effort, according to Biospace (October 05, 2023). The principal role the company was looking to fill 🕹️ was in data engineering, followed by data infrastructure and platform engineering. Other openings included cloud and DevOps engineers, portfolio management and product management, and product managers and facilitators. As of October 2023, about half of those slots have been filled and the new positions had hybrid working arrangements and were based out of London, Seattle, San Francisco and Cambridge, Massachusetts.
More precisely, GSK has brought on Danielle Belgrave as vice president (VC) of AI and ML, that she was most recently at Google DeepMind, creating the health division with the deep learning unit. While the senior vice president and global head of the AI/ML group of GSK is Kim Branson. Shane Lewin is VP in the AI/ML group, Robert Vandersluis is VP of Artificial Intelligence, Jeremy England is Vice President in AL/ML Engineering and Vicki Woodward is Head of AI/ML Operations.
The GSK AI/ML team is focused on the application and development of AI methodology at the intersection of functional genomics and human genetics for target discovery, causal ML and clinical applications. Their main work is to integrate functional genomics, proteomics and transcriptomics data from large industrial-scale experiments and build new models (structured, quantitative data). Plus they have unstructured data—such as notes and reports—for which the AI/ML team is also building natural language models. They operate as a research group and post resources on arXiv and publish results in peer-reviewed journals 📚. Among other things,
They created JulesOS, their proprietary LLM-based operating system.
They built AIML genomic sequence models, particularly those leveraging deep neural networks, that learn complex patterns over long sequences directly from large sequence datasets.
They developed DeepFish, designed to enhance their use of mFISH (Multi-colour Fluorescence In Situ Hybridization) for analysis of Chinese Hamster Ovary (CHO) cell lines expressing therapeutic proteins.
They introduced seq2cells, a ML framework that captures cell-specific gene expression beyond the resolution of pseudo-bulked data and allows variant effect prediction at single-cell level. And
They launched PerturbX, a DL model trained to predict the transcriptional response to a chemical or genetic perturbation in a range of cellular contexts.
At GSK they also have the Onyx Research Data Platform, a suite of next-generation platforms, created to process and scale biomedical data and use it to find new medicines. Their ambition is to build a comprehensive data and ML ecosystem, so that everyone across GSK—especially in Research, AI/ML and Genomic Sciences—has a world-class data experience and the right knowledge at hand when they need it.
On March 30, 2023, PathAI announced its partnership with GSK on HORIZON—a randomized Phase 2b non-alcoholic steatohepatitis (NASH) clinical trial—to measure improvements in liver histology with GSK4532990 compared with placebo in participants with NASH and advanced fibrosis. PathAI’s role will be to generate, digitize, and analyze liver biopsy slides for central pathologist evaluation in addition to AI-powered histologic evaluation using PathAI’s AI-based Measurement of NASH Histology (AIM-NASH) tool.
This collaboration 🖇️ builds on PathAI’s and GSK’s multi-year partnership (announced in 2022) to accelerate scientific research and drug development programs in NASH and oncology.
Boston Massachusetts-based PathAI, provides AI-powered research tools and services for digitizing and analyzing pathology images in order to make safer and more affordable the sub-typing of diseases like breast cancer. PathaAI has the pathology market’s first algorithm to use additive multiple instance learning (aMIL)—the AIM-HER2 Breast Cancer—that delivers automated digital HER2 scoring. MIL models enable spatial credit assignment such that the contribution of each region in the image can be exactly computed and visualized, to provide greater transparency for how AI predictions are made. Apart GSK, PathAI has also collaborations with:
Agilent Technologies Inc. to combine Agilent’s assay development expertise and PathAI’s algorithm development capabilities. By incorporating AI into CDx development, the partnership enables Agilent and PathAI to build integrated solutions with assays and ML analysis algorithms.
ConcertAI to launch a first-in-class quantitative histopathology and curated clinical real-world data (RWD) solution combining PathAI’s PathExplore™ tumor microenvironment panel with ConcertAI’s Patient360™ and RWD360™ products. And
PathAI will work with Roche Tissue Diagnostics (RTD) to develop AI digital pathology algorithms for RTD's companion diagnostics business.
In 2022, 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—expanded its collaboration with GSK to improve clinical trial design, speed up enrolment and identify drug targets. In particular, on October 19, 2022 GSK announced a three-year agreement with Tempus (and a $70M initial payment) that builds on GSK’s and Tempus' existing partnership focused on clinical trial enrolment of patients with specific cancer types.
Tempus provides integrated solutions—sequencing, companion diagnostics, clinical trial solutions, data collaborations, biological modeling etc—by applying AI. Moreover, Tempus introduced in 2023 its standalone RNA next-generation sequencing assay—Tempus xR—a whole transcriptome panel for solid tumors, reporting clinically relevant fusions for more than 100 targeted genes, as well as altered splicing for MET Exon 14 and EGFRvIII. Just a month ago Tempus donated de-identified tumor profiles with limited associated clinical information from over 3,000 cancer diagnoses (from patients sequenced with Tempus’ xT assay), to the National Cancer Institute. In total, Tempus has now raised over $1.3 billion.
In 2022, it was announced that GSK, Sanofi and Takeda have joined an AI and robotics project to increase Singapore's manufacturing capacity for biologics. BioPIPS the Biologics Pharma Innovation Programme Singapore was launched by the Singapore Government to increase the country’s manufacturing capacity for biologics, including recombinant proteins and vaccines. BioPIPS will have three parts:
the “Sensing and Modelling” workstream that will use ML, mechanistic modeling and smart sensors to build simpler and faster workflows,
the “Sustainability” workstream that focuses on tackling sustainability challenges in biologics and vaccine manufacturing, typically using disposable, single-use equipment due to the extremely sterile environment needed for product purity, and
the “Compliant Agility” workstream that focuses on removing manual tasks, using robotics and advanced analytics to increase automation within manufacturing facilities.
GSK, Sanofi and Takeda have joined this project, alongside the National University of Singapore, Nanyang Technological University and the Singapore Institute of Technology.
Regarding Robots, AI/ML and smart manufacturing 🔗, a $120M investment at the Upper Merion, a GSK site in the US in 2019, created a technologically advanced manufacturing hub that offers the flexibility and speed needed to make today’s complex specialty medicines. The strategic investment also laid the foundations for further advances in digitalisation and automation at the site. Other sites across GSK’s network are also implementing technological advances to get ahead. For example, digital mathematical models are enabling teams to predict and improve the performance of vaccines and medicines during the manufacturing process using AI and machine learning capabilities.
Codexis is a leading enzyme engineering company that has a proprietary platform, the CodeEvolver that provides in silico, high-throughput assay screening with AI, and has the power to transform the performance of an enzyme, tailoring 🪡 it for a specific application and process. By using powerful ML tools and sophisticated molecular, cellular and bioanalytical workflows, at Codexis they can design and screen libraries of thousands of enzyme variants in high throughput, then sequence every variant and correlate its sequence with its performance in a highly application-relevant screen. Among Codexis' partners you can find Merck, GSK, Novartis, Nestle, Takeda and many more.
On July 07, 2021, Codexis announced it earned a significant milestone payment related to the licensing of its proprietary CodeEvolver® protein engineering platform technology to GSK.
Codexis has raised 👛 a total of $202M in funding over 4 rounds.
In 2019, Oxford UK-based Exscientia said that it had delivered the first drug candidate to be found using its AI collaboration with GSK—through application of its Centaur Chemist AI-driven drug discovery platform—for a potential therapy for chronic obstructive pulmonary disorder, although the testing timeline remains unclear. Exscientia (Oxford UK, 2012) is the first company to automate drug design and its knowledge-driven systems (Centaur AI, as well as AI Pharmatech platforms) design millions of novel, project-specific compounds and pre-assess each for predicted potency, selectivity, ADME and other key criteria. From this, a selection of the best, information-rich compounds are selected for synthesis and assay. Exscientia has raised a total funding of $429M.
Moreover GSK announced in 2019 to be a member of the project MELLODDY—Machine Learning Ledger Orchestration for Drug Discovery—a landmark project that enabled 10 pharmaceutical companies and seven technology and academic strategic partners to work together and train ML models on datasets from multiple partners while ensuring the privacy of each partner using federated learning.
On July 13, 2022, it was announced that MELLODY has achieved its ambition 🍾 of building a secure platform for privacy-preserving and federated learning. This novel AI framework avoids the need for competitive data and models to ever leave the owner’s custody while still allowing collaborative ML that can train and evaluate drug discovery predictive models.
In 2017, Insilico Medicine established a collaboration with GSK to discover novel biological targets and molecules. In 2020, Insilico also launched a preclinical research program focused on finding new treatments for brain cancer, and has brought on the former global program head of GSK’s computer-aided drug discovery unit to help run it. Insilico Medicine (Rockville, MD USA, 2014) is an AI company that pioneered the applications of the generative adversarial networks (GANs), reinforcement learning, transfer learning and meta-learning for generation of novel molecular structures for the diseases with known and unknown targets, and unlike the other companies in the field is developing the end-to-end pipeline covering every step of drug discovery, clinical trials analysis and digital medicine. Insilico Medicine has raised a total of $401.3M.
And of course it is worth mentioning that GSK was the first 🥇 pharmaceutical company in 2017 to participate in the "Accelerating Therapeutics for Opportunities in Medicine” (ATOM) Consortium, a public-private consortium that aims to cut preclinical cancer drug discovery from six years to just one by leveraging AI. For this project, GSK gave ATOM the chemical and the in vitro biological data for more than 2 million compounds it has screened. The ATOM models and softwares are available on the ATOM Modelling PipeLine (AMPL), an open source, modular, extensible software pipeline for building and sharing models to advance in silico drug discovery.
Finally, on 21 March 2024 a group of researchers from GSK and Imperial College introduced RAmBLA: A Framework for Evaluating the Reliability of LLMs as Assistants in the Biomedical Domain. RAmBLA is a framework proposed to evaluate whether four state-of-the-art foundation LLMs can serve as reliable assistants in the biomedical domain. They identified prompt robustness, high recall and a lack of hallucinations as necessary criteria for this use case. They designed shortform tasks and tasks requiring LLM freeform responses mimicking real-world user interactions. And they evaluated LLM performance using semantic similarity with a ground truth response, through an evaluator LLM. They concluded that larger LLMs generally outperform smaller models across most tasks, particularly in freeform question-answering scenarios relevant to biomedical inquiries, achieving an accuracy of 0.952.
📨 MAbSilico is a TechBio company developing and providing AI-based solutions for antibody drug design and discovery.
MAbSilico was founded in Paris in 2017 by four scientists, whose expertises cover Mathematics, Computational Science, Artificial Intelligence, Structural and Experimental Biology.
On March 18 2024, Rapid Novor Inc. (a world leader in antibody protein sequencing and home to the largest privately funded mass spec proteomics facility in Canada) and MAbSilico announced that they have partnered to provide the world's first AI-driven HDX-MS epitope mapping service for antibody development.
Omnes Capital
Omnes Capital (Paris, France 🥐, 1999) is a leading European private equity firm dedicated to the energy transition and innovation (renewable energy, sustainable cities, deeptech venture capital and co-investment). With over €5.7 billion in assets under management, Omnes provides companies with the equity capital they need in order to grow. Their 55+ employees are based in four offices in Paris, Bruxelles, Munich and Zurich. Marc-Philippe Botte, Michael Pollan and Michel de Lempdes are the managing partners.
Their Venture Capital activity, the historic heart of Omnes with €700M under management, supports innovative European start-ups in the fields of deeptech. Apart the well known AI drug discovery company Iktos they invest into:
DiogenX (TechBio), is a preclinical stage biotech company specializing in the development of pancreatic beta-cell modulators for the treatment of type one diabetes (T1D). Its lead program is a potential first-in-class, disease-modifying recombinant protein designed to modulate the Wnt/β-catenin signaling pathway to regenerate pancreatic insulin-producing beta cells. To date, the company has achieved in vivo 🐁 proof of concept demonstrating efficacy in prevention and reversion settings and an unprecedented effect on human beta cell mass generation, leading to increased functional insulin-producing beta cells. In addition, long term exposure to the drug was well tolerated in all preclinical models and showed the unique ability to safely modulate the Wnt/β-catenin pathway.
The company co-founded by Patrick Collombat, Jean-Pascal Tranié and Benjamin Charles is based in Marseille, France, with research labs in Nice, France, and has raised so far €38M in funding over 2 rounds.
Qantev (SaaS/Cloud) is an AI company that helps health insurers deliver better healthcare to their members. Powered by cutting edge ML and optimization to predict patient journeys, Qantev can provide unprecedented insight and automation by leveraging insurers’ historical claim data. They offer:
Data Foundation technology (ensures the cleanliness, accuracy, and completeness of any health claims data),
Network Management (modern tooling dedicated to health provider performance management, with an automated highlight of pricing gaps and best action recommendation based on advanced simulations),
Claims Management (decision making tool for pre-authorization that enables automated coverage and completeness checks, and advanced quote analysis thanks to medical coding coherence and pricing benchmarks) and
Fraud, Waste & Abuse (detects 🔍complex over-billing and fraudulent patterns, at scale).
The company was cofounded in Paris in 2019 by ML and mathematics experts with background in insurance and finance (Tarik Dadi and Hadrien De March), that were motivated to take breakthroughs in AI and mathematical optimization into the real world. Qantev has raised a total of $11.7M in funding over 2 rounds.
Whitelab Genomics (SaaS/Cloud) was founded in 2019 by David Del Bourgo, MBA—a former engineer at General Electric Healthcare and Julien Cottineau, PhD—a geneticist who trained at the Institut Imagine in France. Together, they developed a simulation software for the design of gene therapies that leverages graph knowledge technology and ML to help scientists discover and design new gene, RNA and cell therapies. On September 13, 2022, Whitelab Genomics raised $10M in series A funding from Debiopharm Innovation Fund, the strategic investment arm of Swiss biopharmaceutical company Debiopharm and Omnes Capital.
In 2023, Sanofi, WhiteLab Genomics, the TaRGeT Laboratory at Nantes University and the Institut Imagine launched the WIDGeT consortium (Viral Vector Intelligent Design for Gene Therapy) to accelerate the development of gene therapies using viral vectors derived from adeno-associated viruses (AAV) for the treatment of age-related macular degeneration and hereditary podocytopathies (kidney diseases), leveraging the potential of AI.
In February 2024, WhiteLab Genomics and Siren Biotechnology, the innovator of universal AAV immuno-gene therapy for cancer, announced a strategic partnership to develop an AI-powered indication discovery platform for AAV immuno-gene therapies for cancer.
Until next time 🌾