Merck KGaA Partners with Insilico Medicine to Deploy Chemistry42™ Generative Chemistry AI Platform

36Kr has learned that the German pharmaceutical giant Merck Group, announced a collaboration with AI-driven drug discovery company InSilico Medicine to integrate its Chemistry42™ generative chemistry AI platform for novel molecular design into Merck Group’s high-performance computing infrastructure（HPC）On the infrastructure front, customized services are provided.

Chemistry42™ is a core component of InSilico Medicine’s Pharma.ai drug discovery platform, integrating artificial intelligence and machine learning methods with specialized expertise in medicine and computational chemistry to design novel small molecules with specific physicochemical properties. The Chemistry42™ platform is a scalable distributed web application capable of synchronously executing multiple tasks within hours. Through container orchestration and workflow management, the platform enables cross-hardware predictive resource allocation and can be deployed on either cloud or on-premises high-performance computing (HPC) infrastructure.

“AI has tremendous potential to transform the drug discovery process, and the flagship product of this generative chemistry AI platform is a testament to that,” said Joern-Peter Halle, Global Head of Research at Merck Group’s Healthcare business sector.

Merck GroupIt is a pharmaceutical giant with over 350 years of history, primarily engaged in pharmaceuticals, life sciences, and chemical businesses. It develops innovative prescription drugs, over-the-counter medications, life science solutions, as well as effect pigments and chemicals for industrial applications.

InSilico Medicine is an AI-driven drug discovery company that has been consistently monitored by 36Kr. It focuses on target discovery, small molecule generation, and clinical trial outcome prediction, leveraging deep generative reinforcement learning technologies. Since 2015, InSilico Medicine has pioneered the global application of generative adversarial networks (GANs) and reinforcement learning (RL).Lai DesignNew drug molecules have led to the publication of over 100 research papers in the field of drug discovery, with more than 25 patent applications filed, including proof-of-concept studies and experimental validation. This year, the company launched the Chemistry42 generative chemistry platform and deployed its services within the first batch of large pharmaceutical companies and drug discovery partner institutions.

Since Ian Goodfellow published his seminal paper on generative adversarial networks (GANs) in 2014, InSilico Medicine has been developing generative chemistry and generative biology algorithms. In 2016, InSilico Medicine publishedFirst Relevant Paper, describes the application of generative adversarial networks in small-molecule drug discovery within the field of oncology. From 2016 to 2020, InSilico Medicine publishedOver 40 papers, and has obtained multiple patents in this field. InSilico Medicine conductedMultiple Proof-of-Concept Experiments, demonstrating that generative models can indeed discover novel targets and design molecules that canEx vivoandIn vivoSynthesis and Testing of Molecules with Specific Properties.

This is a noteworthy phenomenon for players in the field, as leading companies have entered a harvest phase, with an increasing number of collaborations with pharmaceutical firms. This is because, in the AI-driven drug discovery sector, substantial financing alone does not instill sufficient confidence in the industry. Companies and investors are more eager to see validation of the platforms’ AI capabilities, which can be demonstrated either through commercial adoption by pharmaceutical companies or through positive clinical outcomes.

The industry has entered a phase of competition among top-tier players. According to some under-the-radar information obtained by 36Kr, several leading domestic companies have also completed new rounds of financing, with the funds being used to advance their R&D pipelines; this news has not yet been made public.

According to statistics from 36Kr, startups in the medical artificial intelligence sector are concentrated in the niche segments of medical imaging, assisted diagnosis, and disease risk prediction. There were a total of 154 AI projects in the medical field, with over 70% at the angel and Series A funding stages (data as of November 2019).