AI Drug Discovery News (2): Takeda and Sandpiper VC
AI Drug Discovery News (2): Takeda and Sandpiper VC
AI/ML tools, startups, companies and investors
For the first part:
Quote of the Day 🦉
"In God we trust. All others [must] have data. - Bernard Fisher".
By Siddhartha Mukherjee (The Emperor of All Maladies: A Biography of Cancer)
Takeda and AI Drug Discovery
The Takeda Pharmaceutical Company Limited is a Japanese multinational pharmaceutical company. It is the third largest pharmaceutical company in Asia, behind Sinopharm and Shanghai Pharmaceuticals, and one of the top 20 largest pharmaceutical companies in the world by revenue. Its headquarters is located in Chuo-ku, Osaka, and it has an office in Nihonbashi, Chuo, Tokyo. In January 2012, Fortune Magazine ranked the Takeda Oncology Company as one of the 100 best companies to work for in the United States. As of 2015, Christophe Weber was appointed as the CEO and president of Takeda.
On 19 February 2024, Atinary Technologies joined forces with Takeda, for the purpose of merging Atinary’s cutting-edge AI technology with Takeda’s expertise in pharmaceutical research and development, to fast-track the drug development process. Atinary Technologies is a data-driven closed-loop company trying to solve complex optimizations, explore the vast chemical space faster than ever and accelerate innovation and discovery of molecules with the Self-Driving Labs® Platform. SDLabs offers multi-objective optimization (also known as Pareto optimization), that is an area of multiple-criteria decision making that is concerned with mathematical optimization problems involving more than one objective function to be optimized simultaneously.
Founded in Vaud, Switzerland in 2019 by economist-turned-entrepreneur Dr. Hermann Tribukait and chemist and computer scientist Dr. Loïc Roch Atinary has raised so far $5.3M over 3 rounds and its latest funding round was a Seed VC for $5M on September 26, 2023. Atinary Technologies's valuation in October 2022 was $12.3M. The name Atinary comes from the Spanish verb atinar, which means hit, as in hit the target.
Let’s see now Takeda’s AI drug discovery collaborations starting from 2016 🥁
In 2016, Takeda established the R&D Data Science Institute, which integrates data sets such as clinical trials, observational studies, population-level biobanks and real-world data (RWD). Then Takeda used Deloitte’s ConvergeHEALTH Deep Miner™ platform to analyze massive amounts of RWD. Moreover, the AWS Cloud-enabled Deep MinerTM platform used DL and ML techniques to improve the predictive accuracy of Takeda’s earlier results from 53.4% to 92%.
In 2017, Numerate and Takeda announced that they would collaborate on identifying candidates for oncology, gastroenterology and central nervous system disorders. Numerate was acquired by Valo Health (that uses human-centric data and ML anchored computation to transform the drug discovery and development process using the Opal Computational Platform) on November 1, 2019.
In 2017, Recursion Pharmaceuticals (Utah US, 2013) announced the formation of a research collaboration agreement with Takeda for the purpose of providing pre-clinical candidates for Takeda's TAK-celerator™ development pipeline (TAKcelerator, is an in-house semi-autonomous accelerator founded by Tauhid Ali).
🎉 In less than two years, in January 2019, Takeda and Recursion announced progress in their collaboration, reporting that Recursion had evaluated compounds in over 60 indications (rare diseases) during the prior 18 months, which resulted in new therapeutic candidates for more than 6 diseases.
🎉 In May 2020, Recursion Entered Into Global Licensing Agreement with Takeda to Develop TAK-733 in Hereditary Cancer Syndrome. In particular, TAK-733 (REC-4881) is a clinical stage MEK (Mitogen-activated protein kinase kinase also known as MAP2K, MEK, MAPKK) inhibitor, and is being developed to treat a hereditary cancer syndrome and related areas of oncology.
At Recursion Pharmaceuticals’ headquarters clusters of robots treat millions of cells per week with drugs, stain them with six dyes and then take pictures to capture and quantify as many morphological features as they can. By pushing these data through a ML pipeline, they hope to find relationships that are invisible to the human and to tease out clusters of effects that can guide their drug discovery.
In 2019, Image Analysis Group (“IAG”) (London UK, 2007) announced a partnership in the field of ulcerative colitis (UC) histopathological image analysis with Takeda, in order to develop and validate an AI powered tool that will potentially allow gastroenterologists to effectively incorporate histology in their real-life assessment of people suffering from UC. DYNAMIKA is IAG’s cloud platform for robust imaging data management in multi-centre trial setting, and so far has powered over 400 global clinical studies, to manage MRI, CT, X-ray, PET, immunoPET, SPECT, histology and fluoroscopy imaging.
🎇 In 2022, the two extended their partnership to bring AI-powered histopathology image analysis tools into clinics.
In 2020, MIT and Takeda began a partnership within the Abdul Latif Jameel Clinic for ML in Health (J-Clinic), with Takeda providing an initial three-year investment with the potential for a two-year extension to support 6-10 research projects per year, focusing on areas such as disease diagnosis, prediction of treatment response, biomarker development, drug discovery and clinical trial optimisation. Right now, new Takeda fellows are working on electronic health record algorithms, remote sensing data related to environmental health and neural networks for the development of antibiotics.
🧨 In November 2023, the School of Engineering selected 13 new Takeda Fellows for the 2023-24 academic year advancing research at the intersection of AI and health.
🧨 For example, in 2022 it was announced that MIT researchers have developed a geometric DL model called EquiBind that is 1,200 times faster ⏭️ than one of the fastest existing computational molecular docking models, QuickVina2-W, in successfully binding drug-like molecules to proteins. EquiBind is based on its predecessor, EquiDock, which specialises in binding two proteins.