[BioS 레터]’AI role’ throughout the entire process of new drug development

Hyunjin Yang Managing Director, Synteka Bio

Epilogue of Korean-English artificial intelligence symposium

The’Korean-English AI Drug Development and Target Discovery Symposium’ hosted by the Milner Therapeutics Institute in Cambridge, UK and Yonsei University School of Medicine was successfully held online for a total of 3 days from the 22nd to the 24th of last month. The number of symposium enrollees reached 450.

The symposium was held to invite 10 experts in the development of artificial intelligence drugs from both Korea and the United Kingdom to discuss in-depth the current status and prospects of the development of treatments using artificial intelligence. In particular, the discussion focused on how artificial intelligence is being introduced in the process of discovering target proteins and deriving candidates for treatment, which are the initial stages of new drug development. In addition, clinical trials based on artificial intelligence and cases of post-marketing pharmacovigilance were introduced to convey the role of artificial intelligence throughout the entire course of drug development in a balanced manner from various angles.

Professor Jeong Jae-ho of the Yonsei University School of Medicine, who led the symposium, commented on the symposium, “It is the first step to advance the development of next-generation new drugs and to present a new paradigm through academic communication and exchange of research results between a group of artificial intelligence drug development experts between Korea and the UK. Came off. In the future, I hope that it will be an opportunity to further revitalize the exchange of science and technology between the two countries and contribute to solving human diseases and promoting health by jointly cooperating in the development of artificial intelligence technologies related to new drugs.”

AstraZeneca, a multinational pharmaceutical company, and Syntekabio, a domestic artificial intelligence drug development company, and Standigm were also invited as symposium speakers to convey the AI ​​drug development strategy of the new drug development industry.

Universities and research institutes participating in the symposium curate, analyze, and disclose a vast amount of biomedical and chemical research data that is accumulating at a rapid pace with renowned institutions around the world to lay the foundation for the application of artificial intelligence, and to innovate the treatment development process. It is a group of experts who are leading the way. The insights gained from sharing and discussing the vision of AI therapeutics development with them will be an important nourishment for researchers in the future of AI drug development, including Syntekabio.

On the 22nd, the first day of the symposium, it was initiated with a greeting from Tony Kouzarides, head of the Cambridge Milner Institute, and was held by Dr. Rebecca Harris, who served as the head of target discovery at the Milner Institute. ▲Professor Nam Ho-jeong of the Department of Electrical and Computer Engineering at Gwangju Institute of Science and Technology ▲Professor Nam-Sik Han, head of the AI ​​Research Center at Milner Research Institute ▲Dr.

Professor Nam Ho-jeong introduced an artificial intelligence model that predicts the drug-target interaction (DTI) without structural information of the target protein, and the use of this artificial intelligence model is actually large-scale drug screening (High-throughput screening). It has been confirmed that it increases the chances of drug development at the stage. Currently, many AI-based therapeutic candidate screening technologies are based on protein structure information, so Professor Nam’s model is expected to greatly broaden the target protein targets for which AI can be applied. Prof. Namsik Han of Milner Research Institute introduced the process of discovering candidate drugs for re-creation of Corona 19 drugs by using network analysis and artificial intelligence based on the research data of proteomics of the Corona 19 virus along with the introduction of the Milner Institute, which was established for the development of artificial intelligence drugs . In particular, it is characteristic that the changes over time after viral infection are considered together.

Dr. Hee-Jung Koo of Standigm introduces Standigm’s AI drug development platform and covers the early stages of drug development, such as discovering target proteins (Standigm ASKTM), designing candidates (Standigm BESTTM), and searching for new indications (Standigm InsightTM). It emphasized that it has a platform. Next, we demonstrated’iCLUE & ASK’, a target discovery service developed jointly with SK Co., Ltd. C&C, to provide new insights with artificial intelligence to researchers who want to find targets in the new drug development stage. It was conveyed that more personalized analysis is possible by reflecting the user’s idea.

On the second day, on the 23rd, Professor Nam-Sik Han and Dr. Ian Dunham, Director of Open Targets at the European Institute of Bioinformatics (EMBL-EBI), co-chaired. ▲Synteka Biomedical Science Department Hyunjin Yang ▲Director of KAIST Neuroscience-Artificial Intelligence Convergence Research Center, Professor Sangwan Lee, Department of Bio and Brain Engineering, KAIST ▲CEO of Cambridge Crystallographic Data Center (CCDC), a non-profit organization located in Cambridge, UK ) Dr. Juerger Harter presented the presentation.

Syntekabio is discovering candidates for low-molecular compound therapeutics with’DeepMatcherTM’ that simulates the physicochemical properties of the target protein-compound three-dimensional bond structure and predicts it with artificial intelligence, and the’NEOscanTM’ platform Through this, he introduced that he plans to use artificial intelligence for the development of new antigen-based immune anticancer drugs. In particular, he emphasized that he is planning to develop a personalized anticancer immune cell therapy, as he confirmed the excellent performance of Neoscan through experimental verification.

Professor Lee Sang-wan, who is in charge of KAIST Neuroscience-Artificial Intelligence Convergence Center, introduced disease process learning technology that can be applied to the entire healthcare industry, including the development of new drugs tailored to patients, and automatic experiment optimization technology based on disease models. Dr. Yoo Ge-Harter emphasizes the importance of securing high-quality data in order to successfully apply artificial intelligence to new drug research and development. He explained that it is playing a central role as a data center.

On the 24th, the last day of the symposium, Vice Chairman Kathryn Chapman of the Milner Institute took over as chairman. -EBI) Dr. Ian Dunham gave the lecture.

Prof. Yu-Rang Park shared the results of developing an artificial intelligence model related to side effects of anticancer drugs based on data from a multi-center hospital information system in Korea. AstraZeneca Micelle Patel is important in planning and conducting clinical trials, and AstraZeneca also conducts clinical trial feasibility studies, selection of conducting institutions and clinical trial directors, and external control. He said that he is actively using RWD for configuration. However, he stressed that there are tasks to be solved such as integrated analysis of multiple data and bias of data, and advanced statistical analysis techniques such as artificial intelligence are needed to overcome them.

Dr. Ian Dunham, who was the last speaker, introduced the Open Targets Platform developed and operated by the European Institute of Bioinformatics. The open target platform explained the purpose of the development as a tool that enables researchers of new drug development to easily explore potential target proteins related to diseases by integrating a wide range of biomedical data such as disease-related genetic mutation, drug activity data, and literature information.

In the panel discussion following the lecture, major issues related to the application of artificial intelligence were discussed. One of the main topics of the discussion was by far data acquisition. This is because securing high-quality learning data is one of the key requirements for successful AI application. Data usage is frequently restricted due to issues such as personal information protection or data security, and the bias between data according to collection and production methods was also mentioned as a factor limiting the securing of sufficient quantity and quality data for artificial intelligence learning. . In common, the speakers emphasized the importance of sharing data and being open to cooperation between researchers and institutions. In addition, there was also a prospect that the application of technologies such as encryption and blockchain could serve as a solution, as in the example of the’MELODY’ project, a new drug development data sharing platform jointly funded and operated by the European Union (EU) and global pharmaceutical companies.

The importance of cooperation and training of talent were also discussed as major topics. Convergence of various fields such as computer science, data science, and new drug development is important for successful AI drug development, and for this, it is important that experts in each field actively interact with each other within the same goal of developing innovative therapeutics. The need for attitude and communication was emphasized.

Furthermore, an opinion was shared that changes in the curriculum are also necessary to cultivate talents suitable for convergence research. Despite the short history of AI drug development, which is less than 10 years old, there have been continuous advancement of technology and gradual achievements, and as these achievements are now becoming increasingly visible, artificial intelligence will be able to contribute to the development of therapeutics at a faster pace. The symposium ended with hopeful prospects.

Meanwhile, this symposium is jointly held by Yonsei University and Cambridge University as part of the Korea-UK Science and Technology Cooperation Center (Focal Point) project supported by the Ministry of Science and Technology Information and Communication to promote researcher exchange activities and science and technology cooperation between Korea and the United Kingdom. I did.

The Milner Institute, which co-hosted the symposium, is a research institute affiliated with the University of Cambridge University School of Medicine and was established in 2015 to contribute to the development of new treatments using new research methodologies such as open artificial intelligence. Europe’s largest open innovation bio drug development cluster in which 85 companies, universities, and research institutes around the world, including global pharmaceutical companies such as AstraZeneca, Johnson & Johnson, and GSK, cooperate with each other for close industry-academia-research cooperation research. Is leading the way. In Korea, Samsung Medical Center, Yonsei University Severance Hospital, KAIST, Korea Electronics and Telecommunications Research Institute (ETRI), and JW Pharmaceuticals have joined as consortium members.