AI/ML tools, startups, investors and companies transforming drug discovery 🪁
A newsletter about Science, Technology and AI Drug Discovery
“Our brains contain one hundred billion nerve cells (neurons). Each neuron makes links with ten thousand other neurons to form an incredible three dimensional grid. This grid therefore contains a thousand trillion connections - that's 1,000,000,000,000,000 (a quadrillion). It's hard to imagine this, so let's visualise each connection as a disc that's 1mm thick. Stack up the quadrillion discs on top of each other and they will reach the sun (which is ninety-three million miles from the earth) and back, three times over.”
By Nessa Carey, The Epigenetics Revolution
News
⚽️ AION Labs launches CombinAble.AI to design antibodies for therapeutic development
🏀 Relation Therapeutics lands $35M to discover osteoporosis drugs with AI
Merck and AI Drug Discovery
Merck, a leading science and technology company, apart launching its AIDDISON™ drug discovery software—the first software-as-a-service platform that bridges the gap between virtual molecule design and real-world manufacturability through SynthiaTM retrosynthesis software application programing interface (API) integration—has made so far a significant number of partnerships with AI drug discovery companies/startups, combining generative AI, ML and computer-aided drug design in order to increase the success rate of new drugs and therapies. For example:
🔶 On January 25, 2024, Variational AI announced a project with Merck Research Labs supported by the CQDM Quantum Leap program. Variational AI in Vancouver leverages a powerful new form of ML known as generative AI to free scientists from reliance on screening and libraries (both experimental and virtual) and to eventually generate de novo molecules with all the optimized properties, in order to discover efficacious, safe and synthesizable small molecule therapeutics in a fraction of the time and cost.
Basically, their answer is not to screen molecules, but to create molecules with a platform called Enki. The Enki algorithm learns from a training set of molecules screened against drug targets (from experimental and computational sources) and then generates novel molecular structures with multiple predefined parameters optimized to avoid common causes of drug attrition. Variational AI has raised $3.7M.
🔶 On November 29, 2023, Amazon Web Services, Inc. (AWS) announced it is working with Accenture to help Merck—known as MSD outside of the U.S. and Canada—to move a substantial portion of its IT infrastructure to AWS as part of the company’s multiyear cloud migration. As part of this initiative, Merck selected AWS as its preferred cloud services provider and Accenture as its professional services partner to transition core applications like SAP, ML and data warehouses to AWS—enabling Merck to help accelerate scientific research and discovery.
🔶 On September 20, 2023, Merck KGaA announced three-target deals with BenevolentAI and Exscientia. The deals are collectively worth more than $30M upfront and $1 billion in potential milestones. BenevolentAI and Exscientia will work on small molecules for oncology, neuroinflammation and immunology using their AI platforms to provide candidates that the German drugmaker can move through preclinical development and into the clinic.
BenevolentAI (UK, 2013) is using AI for scientific data mining ⛏️, data contextualisation and deriving hypotheses. In particular, they have developed the Benevolent Platform® (the platform is disease agnostic) built on powerful data foundations, which has the ability to rapidly normalize and contextualize a diverse range of external and internal data from numerous sources including, but not limited to, scientific literature, patents, genetics, chemistry and clinical trials. BenevolentAI has raised a total of $292M.
Exscientia Ltd (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.
In 2020, Exscientia announced that the first-ever AI-designed drug molecule (DSP-1181) in collaboration with Sumitomo Dainippon Pharma, had entered human clinical trials. DSP-1181 was developed in only 12 months (instead of 5 years the average “ex normal”), in order to treat obsessive-compulsive disorder. Moreover, EXS-21546 (for solid tumors carrying high adenosine signatures) by Exscientia and Evotec and EXS-4318 (for inflammatory and autoimmune conditions) by Exscientia in collaboration with Bristol Myers Squibb are already in or are entering clinical trials.
🔶 JW Pharmaceutical announced at the beginning of 2023 that it has signed a business agreement with Merck of Germany at its headquarters in Seocho-dong, a district of the South Korean city of Seoul. Under this agreement, Merck will provide Merck's AI Synthia retrosynthesis software to JW Pharmaceutical. Synthia harnesses the power of AI in chemistry to curate information and predict reactions for computer-aided synthetic design of optimal synthetic routes.
🔶 On January 17, 2023, Merck selected Perceiv AI for Inaugural Digital Sciences Studio Cohort. Merck Digital Sciences Studio (MDSS), is a collaboration between Merck and the New Jersey Innovation Institute, with investments from the Merck Global Health Innovation Fund, Northpond Ventures and McKesson Ventures. The accelerator provides a direct investment from Merck and also access to the Azure Cloud computing and technology support provided by Microsoft for Startups. Perceiv AI is an AI driven precision medicine company developing a multimodal prognostic platform to forecast disease progression in age-related disorders. Perceiv (Canada, 2018) has raised a total funding of $1.5M over 1 round.
🔶 Merck selected in 2022 Saama to add ML tech to drug development process. Saama is a data analytics company and will work with Merck & Co. to build a new clinical data management system, using its own Life Science Analytics Cloud (LSAC). LSAC is powered by cloud computing and AI/ML technologies, trained using more than 100 million clinical data points. Saama’s system is technology-agnostic, so it can be easily slotted into Merck's existing infrastructure. Saama (US, 1997) has raised a total funding of $75.8M over 3 rounds.
🔶 Also in 2022, Merck selected Numerate for computerized cardio drug discovery. Numerate is a technology platform company that is leveraging proprietary algorithms and the power of cloud computing to transform the drug design process. Numerate (US, 2007) has raised a total of $19.4M in funding over 5 rounds.
🔶 In 2021, Syapse announced that it has entered into a research collaboration with Merck (Source), known as MSD outside the United States and Canada. Syapse develops software for aggregating and structuring biomedical data. In particular, powered by Syapse’s Raydar, their platform seamlessly blends data intelligence, insight analytics and applied experience through healthcare organizations looking to transform patients’ lives using real-world evidence. In this way, they transform highly contextual, clinically complex oncology data into actionable insights. In 2022, Syapse (US, 2008) raised $35M in growth funding round, for a total of $222.6M in funding over 9 rounds.
🔶 In 2021, Merck announced that it will work with McKesson's Ontada to conduct real-world evidence studies, since Ontada has data from more than 2 million records and will allow Merck to explore new scientific data to improve patient outcomes and the quality of its cancer therapies. Ontada already works with the US Oncology Network, which taps into 1,400 independent physicians for value-based care.
🔶 On June 25, 2021, Nucleai—a precision medicine company using AI-powered image analysis technology to support novel pathology-based biomarker development—announced that it has partnered with Merck KGaA, Darmstadt, Germany, in order to leverage Nucleai's image analysis and biomarker discovery platform for several Merck's clinical stage oncology assets. In 2022, Nucleai announced the expansion of its spatial biology platform to include a new generation of multiplex immunofluorescence analysis that uses DL to further unlock the power of multiplex immunofluorescence data analysis during drug discovery and development. Nucleai (Israel, 2018) has raised a total of $46.5M.
🔶 In 2020, Insilico Medicine announced that Merck KGaA, Darmstadt, Germany will be the first launch partner for its flagship generative chemistry AI platform Chemistry42, and that Merck will integrate Chemistry42 into its discovery pipeline to facilitate rapid and effective drug design. Chemistry42, core part of Insilico's Pharma AI drug discovery process, is a user-friendly platform that bridges AI and ML methods with domain expertise in the fields of medicinal and computational chemistry, for the design of novel small molecules with desirable physicochemical properties. Insilico Medicine (Hong Kong and New York, 2014) has raised a total of $401.3M.
🔶 In 2020, Iktos—a company utilizing AI for new drug design based on deep generative models—announced a collaboration agreement with Merck KGaA, Darmstadt, Germany. In this new collaboration, Iktos’ de novo generative design technology will be used in a structure-enabled context to facilitate the rapid and cost-effective design of an undisclosed Merck drug discovery program. Iktos has developed so far: DockAI (docking), Makya (de novo design), Spaya (retrosynthesis) and SpayaAPI (a scoring tool for virtual molecule libraries). Iktos (Paris, Ile-de-France, France, 2016) has raised a total of $15.5M.
🔶 In 2019, BioXcel Therapeutics—a clinical stage biopharmaceutical company utilizing AI approaches for neuroscience and immuno-oncology drug development—announced the addition of Merck KGaA, Darmstadt, Germany to its Nektar Therapeutics clinical collaboration to evaluate a novel triple combination therapy in pancreatic cancer (avelumab, BXCL701 and NKTR-214). In 2023, BioXcel reported promising top-line data including an “encouraging” response rate in its phase combination clinical trial. BioXcel Therapeutics aims to become the leading AI-enabled neuroscience company, and entered a $260M long-term financing agreement with Oaktree Capital Management and Qatar Investment Authority to accelerate growth and commercially launch a new schizophrenia drug (IGALMI™ dexmedetomidine sublingual film) in return for a share of the profits. BioXcel Therapeutics (US, 2017) has raised a total of $140M in funding over 2.
🔶 In November 2019, Boston-based PathAI—maker of AI tools for pathology—capped off its Series B funding round with a $15M strategic investment from the Merck Global Health Innovation Fund and Bristol-Myers Squibb. 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. For example, 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. PathAI has raised a total of $355.2M in funding over 6 rounds.
🔶 In 2018, Merck KGaA revealed that has entered into a licensing deal with Cyclica—a Toronto-based biotechnology company that leverages biophysics and AI to make drug discovery faster, safer and cheaper—in order to utilize Ligand Express a cloud-based proteome screening platform, that screens small-molecule drugs against repositories of structurally-characterized proteins or ‘proteomes’ to determine polypharmacological profiles.
Cyclica has two ML engines underlying its platform: MatchMaker™ and POEM™.
MatchMaker, is Cyclica’s core differentiator, an AI-enabled DL engine that predicts the polypharmacology of small molecules as the foundation for small molecule drug discovery. It is able to generalize across the proteome and uses both AlphaFold2 structures and homology models.
And POEM (Pareto Optimal Embedding Model) is a unique similarity-based property prediction model. In contrast to other AI prediction models, POEM uses multiple types of molecular fingerprints to describe molecules, providing a much richer measure of similarity that leads to greater accuracy.
Cyclica and valence_ai were acquired by RecursionPharma—a leading clinical stage TechBio company fuelled by 19 petabytes of data and one of the world’s most powerful supercomputers—on May 8, 2023.
🔶 Moreover, Merck has a confidential project (2018) with Atomwise—a company that utilizes AI, specifically convolution neural networks, in order to assist with drug discovery efforts. Atomwise (San Francisco US, 2012) is a pioneer of harnessing AI for drug discovery and invented the first DL technology for structure-based small molecule drug discovery, and holds several patents for applying convolutional networks to spatial data. Atomwise in 2020 scored a whopping $123M in Series B funding led by B Capital Group and Sanabil Investments. DCVC, BV, Tencent, Y Combinator, Dolby Family Venture and AME Cloud Ventures also participated in the funding round. The company cut its headcount by 30% in December 2022, just months after its AI deal with Sanofi. So far, Atomwise has raised $176.6M.
🗣️ A look into RWKV: A more efficient answer to Transformers? [Breakdowns]
By
from Artificial Intelligence Made SimpleSiftMed
SiftMed in Canada 🍁 is an AI tool that is designed to help organize and analyze medical files for personal injury practices. SiftMed has revolutionized medical document management with its specialized AI investigative platform, tailored to meticulously organize and categorize medical files, and significantly cut ✂️ review time by 50% compared to traditional manual methods. With SiftMed you can drag and drop files from your case management system or directly from your computer, that automatically detects duplicates, creates a chronology, categorizes records by record type, and extracts handwritten text. Moreover, it can easily flag and tag events, make notes and spend less time searching for keywords, and finally build your final report, export and move onto the next case.
Founded in 2020 by Holly Hill, CEO & Co-Founder and Jeremie LaRouche, M.D., SiftMed has grown significantly and the team is now a diverse group of innovative leaders, data scientists, software developers and medical experts. In 2023, SiftMed raised a $2.7M Seed Round led by Sandpiper Ventures.
Something funny 😋 from LinkedIn
Blume VC
Blume VC is backing the next wave of revolutionary founders from India 👳. Blume focuses on Seed to Pre-Series A stages of funding and has made over 100 investments across its 11-year history across multiple funds. In 2010, Karthik Reddy and Sanjay Nath founded Blume Ventures with a mission to reimagine startup financing for India. Sanjay also represents Blume on the boards of Arka Venture Labs (an India-US cross-border accelerator platform, launched in partnership with three leading VC firms: Benhamou Global Ventures/BGV, Blume Ventures and Emergent Ventures) and the Draper Venture Network (a robust venture collective that brings together experienced investors and innovative entrepreneurs from across sectors and continents to share strategies, source opportunities and create greater value).
Blume invests into:
BeatO (2015) offers a full stack solution comprising products and services specific to diabetes and other chronic care needs. These products and services include glucometers as well as comprehensive care programs to prevent, control & reverse Diabetes and other chronic diseases.
Karmic Lifesciences is a full-service contract research organization headquartered in India with global execution capabilities in the US, Europe and APAC. It provides clinical trials and clinical data management & bio-statistics projects (Phase I to IV) in various therapeutic areas including Oncology, Cardiovascular, Diabetes, Neurology, Auto-Immune Diseases, Ophthalmology, and Medical Devices. It also launched medico-marketing services for the pharmaceutical industry, and developed kTrials, a web-enabled eClinical suite with integrated clinical trial management, clinical data management, clinical data warehousing and operational analytics functionalities. It was funded by India Angels Network, Mumbai Angels, Basil Growth, Blume Ventures. It was acquired by Cliantha Research in July 2014.
THB (2015) stands for Technology | Healthcare | Big Data Analytics, and is amongst India’s leading clinical intelligence companies. THB leverages real world evidence with state-of-the-art clinical and business modules to improve patient outcomes, thereby aiming to move the healthcare industry from gut-based decisions to data-driven, evidence-based treatment and decisions.
Tricog’s (2014) cloud-connected InstaECG devices, located at local clinics, help detect heart complications within minutes of a patient’s arrival. Once detected, Tricog facilitates care coordination between the remote location and a tertiary center that would treat the patient so that even before the patient reaches the hospital, an expert team is ready to treat the heart attack without delay – thereby saving lives.
Ultrahuman (2020) helps the health-conscious track and enhance their metabolic fitness through devices (continuous glucose monitoring patch, the ring) as well as an app that has diet and regimen recommendations.
For the second part ➡️: AI/ML tools, startups, investors and companies transforming drug discovery (2nd part). 🪁
HI Mariana ... I enjoyed the articles. There is one big clarification on the "Merck and AI Drug Discovery " article: There are two different Merck companies - US and Germany ... the article mix news from both as the same Merck ... I rectification will be appreciate it. Thanks and keep the good work.