[Reading Science] "Interpreting Every Single Cell"... Creating a Human 'Digital Twin' with AI

The era of replicating the human body as "data" is becoming a reality. The concept of a "digital twin," which reads the state of each individual cell and integrates this information with artificial intelligence (AI) to predict diseases and immune responses, is emerging as a new research trend in biology.

On April 27, at the "Science Journalists Association-IBS Science Media Academy" held in the Premium Conference Room on the 2nd floor of the Seoul Station Biz Hub Center, Jong-eun Park, Director of the IBS Virus Research Center and Professor at the Graduate School of Medical Science and Engineering at KAIST, gave a lecture on "The Mysteries of Our Body and Diseases Realized through Cellular Maps and Artificial Intelligence."

Jong-eun Park, Director of the IBS Virus Research Center and Professor at the Graduate School of Medical Science and Engineering at KAIST, is giving a lecture on "The Mysteries of Our Body and Diseases Realized through Cellular Maps and Artificial Intelligence" at the "Science Journalists Association-IBS Science Media Academy." Photo by Kim Jounghwa

원본보기 아이콘

Director Park introduced a research direction that reconstructs human biology as a "simulatable system" by combining single-cell analysis, spatial omics, and artificial intelligence (AI). He said, "Now, the human cell map is almost complete," adding, "To truly explain the phenomena of life, we need to understand how cells respond when stimulated."

The human body consists of approximately 30 trillion cells. These cells share the same genome, but they perform different functions depending on which genes are expressed.

The challenge is that it is virtually impossible to understand such vast information simultaneously. Director Park compared this to "a massive matrix entangling 30 trillion cells and tens of thousands of genes," explaining, "The key is to compress this information into a structure that can be understood."

To achieve this, researchers are building an AI-based "foundation model" that comprehensively reflects gene regulation within cells, intercellular interactions, and organ-to-organ structural connections.

Director Park emphasized, "It is important to quantitatively understand how genes cooperate and inhibit each other," adding, "Modeling these networks is essential to implementing a digital twin."

This approach is now expanding its scope in actual research fields. For example, studies on "blood phenotype," which distinguish disease states based on the composition of cells in the blood, are underway. These efforts aim to classify a range of diseases such as autoimmune and infectious diseases and to predict drug responsiveness.

Director Park explained, "By classifying patients in advance based on their cellular state and linking this with drug responsiveness, we can predict therapeutic effects," adding, "Such approaches are already in use in the field of precision medicine."

A conceptual diagram showing the process of creating a "virtual human (digital twin)" by analyzing single-cell and spatial omics data with artificial intelligence, then applying stimuli to predict disease responses. Provided by IBS

원본보기 아이콘

Additionally, research is being conducted to elucidate the mechanism of action at the cellular level, such as confirming through single-cell analysis that mRNA vaccines selectively act on specific cells and induce immune responses.

However, this technology is not yet at a stage where it can completely replace traditional experiments. He stated, "Current models excel at understanding at the cellular level, but animal experiments are still necessary to study whole biological systems where multiple tissues interact simultaneously," predicting that "it will take at least five to ten more years for complete replacement."

This research trend is also interpreted as an example of the changing nature of biological research methods. While traditional biology focused on observing and analyzing individual phenomena, there is now a growing effort to understand intercellular interactions and disease progression processes comprehensively, based on large-scale data.

Director Park said, "The important thing is the process of generating and verifying hypotheses through data," adding, "As more data is accumulated, the ways we understand and predict diseases will become increasingly sophisticated."

He further commented, "Although the digital twin encompasses various concepts, in the end, it is an attempt to quantitatively understand the relationships between cells and genes," adding, "These models are laying the foundation for a more systematic explanation of human biology."

© The Asia Business Daily(www.asiae.co.kr). All rights reserved.