pharma-Nearly 80 AI Startups Created in China’s Ph

Credit: Visual China

BEIJING, December 14 (TMTPOST) – The Global Health Drug Discovery Institute (GHDDI) announced a collaboration with Microsoft Research AI4Science on Wednesday, with an aim to develop generative artificial intelligence (AI) and basic large-model technologies in the field of infectious diseases to speed up drug discovery and development.

The GHDDI, initiated by Tsinghua University, the Bill & Melinda Gates Foundation, and the Beijing Municipal Government, operates as an independent, non-profit global public health and drug innovation institution. Its mission is to facilitate more efficient drug discovery and development to address global health threats.

According to information obtained from a recent meeting held by GHDDI, from 2010 to 2021, the number of drug development projects involving AI technology had grown from 6 to 158 globally.

Traditional computer-aided drug design relies on physics, while AI-assisted drug development primarily involves data-driven deep learning and large-model technologies, making it simpler and more versatile. AI algorithms have been able to empower drug discovery and development in various stages, including target discovery, drug repurposing, compound screening, molecular design and optimization, and patient recruitment grouping.

GHDDI's collaboration with Microsoft Research AI4Science focuses on infectious diseases threatening global health, such as tuberculosis, viral infections, and malaria, integrating generative AI, basic large-model technologies, virtual high-throughput screening, and expert experience. Leveraging GHDDI's drug development capabilities and its platform dedicated to turning cutting-edge technologies into products, the collaboration aims to achieve a complete process from virtual compound generation and screening to synthesis and optimization, biological experimental verification, and then to AI iterative upgrading.

Zhang Rumin, the Chief Scientist of the GHDDI, said that compared to the lengthy traditional drug development process, AI technologies can successfully design small molecule inhibitors with significant effects against relevant pathogens within 3 to 5 months. For example, in the research on Mycobacterium tuberculosis and coronaviruses, the rapid design of new small molecule inhibitors demonstrated the potential of AI in drug design.

In studies targeting Mycobacterium tuberculosis and coronaviruses, the joint team of GHDDI and Microsoft Research AI4Science has successfully designed various effective small molecule inhibitors against key target proteins of these pathogens through the innovative generative AI drug development process in just six months. These compounds performed exceptionally well in enzyme inhibition experiments, with inhibitory efficacy even surpassing known lead compounds.

Guo Jinjiang, Head of Data Science at the GHDDI, said that AI technology has a significant impact on the field of life and health. Especially in the drug development model where AI algorithms are used to repurpose existing drugs, there can be savings of 3%-50% in both time and cost. However, he cautioned that although AI technology can accelerate the drug development process and reduce costs, there are many challenges in the entire process. These include how to better enable AI to understand and learn biomedical data, and how to improve collaboration between biopharmaceutical companies and technology giants in the implementation and transformation of AI.

Additionally, experts said at the recent conference of the GHDDI that AI has brought about significant disruptions to many industries, and there are challenges in terms of ethics, policies, and values. Authorities need to formulate policies and regulations, and establish a relatively complete mechanism. Clearer regulatory boundaries are required to further promote the development of AI technology in a direction beneficial to humanity.

As of the end of 2022, there were over 700 AI pharmaceutical companies globally, and nearly 80 in China.

The drug discovery and development process often takes a decade and requires an investment of $1-3 billion, with a success rate of nearly 10%. With the emergence of AI-powered pharmaceutical startups, there is the potential for transformation in the pharma industry.

(This article was originally published on the TMTPost App. Author: Lin Zhijia.)