Global Health Drug Discovery Institute (GHDDI) (Opens in new tab) Microsoft Research has recently made significant progress in accelerating drug discovery for the treatment of global infectious diseases. Working closely together, the collaborative team successfully used generative AI and fundamental models to design several small molecule inhibitors against essential target proteins of Mycobacterium tuberculosis and coronaviruses. These new inhibitors exhibit superior biological activity comparable to or exceeding the best known lead compounds.
This breakthrough is a testament to the team’s collaborative efforts in generative AI, molecular physical-chemical modeling, and iterative feedback loops between scientists and AI technologies.Usually discovered and in vitro Confirming such molecules could have taken up to several years, but with the acceleration of AI, the collaborative team achieved these new results in just five months. This research also highlights the huge potential of AI to help scientists discover or create the building blocks needed to develop effective treatments for infectious diseases that continue to threaten the health and lives of people around the world. is showing.
For example, since 2019, there have been more than 772 million confirmed cases of COVID-19 worldwide, according to the World Health Organization (WHO), the Centers for Disease Control, and various other sources. The death toll has reached nearly 7 million people. Although vaccines have reduced the incidence and duration of the disease, coronaviruses continue to mutate and evolve, posing a continuing and serious threat to global health. Meanwhile, according to a WHO report, tuberculosis remains the leading cause of death among infectious diseases, and will be the second leading cause of death after the new coronavirus infection in 2022, with 10.6 million people worldwide suffering from tuberculosis. , 1.3 million people died (latest figures currently available).
Laying the foundation for new infectious disease treatments
Microsoft Research has extensive experience in developing and pre-training large-scale protein and molecule-specific AI models, demonstrated in both property prediction and molecule generation. Based on these experiences, Microsoft Research has developed and maintains ownership of AI models for the generation of molecules tailored to specific protein targets. The generated compounds were virtually screened and further optimized by GHDDI data scientists and medicinal chemists, followed by compound synthesis and wet lab experiments to quantify bioactivity. The experimental results were fed back to Microsoft’s research team to improve the AI model and generate new compounds.
This integrated AI, expert and experimental pipeline successfully generates novel compounds against protein targets of Mycobacterium tuberculosis and the coronavirus SARS-CoV-2. Within five months, the collaborative team designed several compounds effective in inhibiting essential target proteins of these pathogens, accelerating the structure-based drug discovery process.
One of the distinguishing features of the AI-generated molecules is their novel scaffold structure. This is important because it creates the potential for these molecules to be developed into new classes of drug candidates. These new structures not only offer the potential for more effective treatments, but also address the escalating challenge of antimicrobial resistance (AMR), which is a major hurdle in the treatment of infectious diseases such as tuberculosis and COVID-19. It also helps in dealing with.
“In the current scientific research landscape, we face unprecedented challenges, but also unprecedented opportunities,” said Dr. Shen Ding, Director of the GHDDI Institute. “Innovation is a central catalyst for scientific progress and a key element in addressing global health challenges. I am excited about our collaboration with Microsoft Research and am pleased with the progress we have made together. We have no doubt that our combined efforts will improve research and development efficiency and speed the drug discovery process.”
“This represents collaboration across disciplines and boundaries,” he said. “Together, we will advance pharmaceutical research and open new avenues of scientific exploration.We look forward to seeing cutting-edge technologies like this deployed into uncharted territory in the life sciences. This will enable us to provide more comprehensive, deep and practical solutions to global health challenges.”
Spotlight: Microsoft Research Newsletter
Microsoft Research Newsletter
Stay connected to Microsoft’s research community.
Improving global health with AI
Adopting open innovation principles, the GHDDI and Microsoft Research collaboration is dedicated to leveraging AI technology to accelerate drug discovery. Our goal is to contribute to global health equity through the development of lifesaving medicines and the rapid delivery of safer, more effective pharmaceutical solutions that are accessible to everyone. This collaboration focuses on infectious diseases that pose a threat to global health, including but not limited to tuberculosis, viral infections, and malaria. Both parties are working on generative AI, foundational models, high-throughput virtual screening, and tight integration of expertise to address these challenges.
“Successful AI-enabled drug discovery requires close collaboration between AI experts and medical experts,” said Dr. Tie-Yan Liu, Distinguished Scientist at Microsoft Research AI4Science. says. “In recent years, Microsoft Research’s globally recognized team has been deeply engaged in interdisciplinary research between AI and the natural sciences. To complement this, GHDDI’s experts bring extensive industry experience and They bring deep domain knowledge. Their laboratory facilities not only enable testing, but also help provide valuable feedback to train the AI models. We work closely together. We look forward to producing groundbreaking research findings that have the potential to redefine the future of healthcare through innovations in AI technology.”
Acceleration of drug discovery
Commenting on research into Mycobacterium tuberculosis and coronaviruses, GHDDI Chief Scientific Officer Dr. Ruming Zhang said that the application of AI technology by the collaborative team was able to significantly shorten the traditional lengthy drug discovery process. said. The team was able to design and validate a highly effective small molecule inhibitor against the pathogen in just five months.
“This is an extraordinary achievement that highlights the immense potential of efficient AI. de novo medicine design. It also vividly demonstrates the team’s exceptional innovative capabilities and professional capabilities,” he said. “We are excited that this innovative R&D strategy will lead to more breakthrough advances in a broader field of future drug discovery projects.”
Dr. Tao Qin, Senior Principal Scientist at Microsoft Research AI4Science said: “This effort is about pushing the limits of AI technology that can be applied to new drug research and development. We will ensure that the benefits are available to everyone.”
Dr. Jinjiang Guo, Head of Data Science Department at GHDDI, said, “We plan to strengthen and expand our cooperation and further promote the use of AI technology in the field of life sciences.” “This provides new insights that enrich researchers’ understanding of the mechanisms underlying disease and life, and paves the way for the development of innovative therapeutic strategies that have long impacted human health. We are very optimistic about the potential of this collaboration and believe it will have a significant impact on the future of the healthcare sector. ”
next step
In the next phase, Microsoft Research and GHDDI will collaborate to optimize the discovered hits, enhance the properties of ADMET, advance toward preclinical studies, and initiate broader drug discovery projects.
