A Peek into the Future of Data and Science

Researchers and scientists are always looking for new ways to uncover and answer nature's biggest questions. On November 14, the Data Innovation in Science Symposium showcased recent findings made in fields such as chemistry, biology, and earth sciences by Seoul National University researchers, in collaboration with international researchers. The symposium explored the clever usage of artificial intelligence and big data in furthering the researchers' respective fields.

Poster for the Data Innovation in Science Symposium

Professor Lee Jae Kyoo from the Department of Applied Bioengineering took the stage to present his team's research, titled "AI-Driven Optimization of Energy-Conversion Nanomaterial Catalysts." This project delved into microdroplet chemistry and was conducted alongside Professor Hwang Yun Jeong from the Department of Chemistry and Professor Fritz Printz from Stanford University. Microdroplets, Professor Lee explained, are tiny droplets of water smaller than the width of a human hair. These are particularly interesting due to their unique properties: a strong intrinsic electric field and the ability to accept or donate electrons, which creates a highly reactive environment. As a result, microdroplets serve as excellent catalysts, especially compared to reactions taking place in regular test tubes.

Professor Lee and the team are exploiting these properties to create nanoparticles, or high-performance catalysts. Compared to other methods currently used, which are often time-consuming, energy-intensive, and require toxic chemicals, the microdroplet method only uses water to synthesize high-quality crystals. Not only does it prove to be a more environmentally friendly process, it also allows for faster and cheaper production of these vital components used in everything from fuel cells to hydrogen production.

Professor Lee presenting "AI-Driven Optimization of Energy-Conversion Nanomaterial Catalysts"

In order to control the complex conditions within microdroplets, such as size, pH, and temperature, the team developed a specialized AI model. This AI model learns from large public reaction databases and is then fine-tuned with the data collected from the microdroplet experiments. What would have been a tedious trial-and-error process is streamlined as AI predicts the optimal conditions for the best results.

Diving into recent developments in earth sciences, Professor Woo Jusun from SNU's Department of Earth and Environmental Sciences delivered a presentation titled "Co-evolution of Planets and Microbes." He focused on the symbiotic relationship between our planet's geology and the evolution of life. For example, he highlighted the way the roots of small plants were able to shift river systems into the meandering rivers we know today by stabilizing the soil. More broadly, Professor Woo had three key research areas: changes in microbial sedimentary structures over time, sedimentary records during the Late Paleozoic, and geological-microbiological interactions in mine drainage. To achieve this, he worked with the Deep-Time Digital Earth (DDE) Program.

Professor Woo presenting "Co-evolution of Planets and Microbes"

Professor Fan Junxuan from Nanjing University gave a detailed explanation of the program. The DDE Program is a large international science program that aims to link distributed systems of global Earth evolution data, transforming earth science into a data-intensive discipline. So far, the team behind the program has compiled a database with 2.2 million fossil records, making it the world's largest in sedimentary paleontology. Now, the DDE team has created a new initiative called the Global Open Science Fund (GOSF), hoping to attract diverse international investments and support major research projects.

The symposium wrapped up with talks from graduate and doctoral students. Choi Min-Soo from the School of Biological Sciences spoke on his research using AI to verify the functions of chemical compounds. Prior to this, the process could take over a month of purification and testing, which made testing large numbers of compounds virtually impossible. Using advanced statistical modeling AI as a solution, Choi has been able to test mixtures of compounds rather than examining them individually.

The Data Innovation in Science Symposium proved to be a fruitful time for researchers to gather and share their recent achievements in their respective fields, especially presenting how technologies such as AI are being wielded to further scientific research. The findings promise many interesting breakthroughs in the future, leaving all excited for news of more.

Written by Lee Eusun, SNU English Editor, [email protected]