'AI-enabled stem cell for pediatric epilepsy proves effect'

100 times greater therapeutic effect than existing treatments

KHIDI-Korean researchers discover personalized new drug candidate

Personalized stem cell modeling and AI-based drug discovery technology

A new treatment possibility has opened up for pediatric epilepsy patients who have shown little response to existing treatments.

 

The Korea Health Industry Development Institute (President: Soondo Cha) announced on the 19th that Professor Hoon-Chul Kang’s research team at Yonsei University Severance Children's Hospital has successfully discovered a new drug candidate personalized for patients and experimentally proven its efficacy by utilizing AI-based drug discovery technology and induced pluripotent stem cells (iPSCs) obtained from patients.

 

Pediatric epilepsy is a representative intractable neurological disorder, affecting approximately 250,000 people in Korea, with about 30-40% of patients suffering from drug-resistant epilepsy that does not respond to existing antiepileptic drugs.

 

(from the left) Professor Hoon-Chul Kang and Ji-hoon Kim from Yonsei University and Professor Do Kyun Na from Chung Ang University

In particular, pediatric epilepsy patients with rare genetic mutations, such as SCN2A gene mutations, exhibit significant differences in experimental responses by individual patient, rendering it difficult to achieve significant improvements with existing treatments.

 

In addition, the lack of precise disease models and appropriate drug screening technologies has made the development of personalized drugs an urgent necessity.

 

A joint research team led by Professor Hoon-Chul Kang and Ji-hoon Kim from Yonsei University and Professor Do Kyun Na from Chung Ang University addressed this issue by creating induced pluripotent stem cells (iPSCs) from patients' blood cells to develop precision disease models in the same disease environment as actual patients.

 

Also, the team successfully identified new drug candidates, surpassing existing treatments through AI-driven high-throughput compound screening and validation.

 

The research team used the latest gene editing technology to correct the SCN2A mutation to its normal state to reveal that the mutation is the direct cause of epilepsy, and confirmed that seizure symptoms disappeared.

 

Furthermore, based on personalized neural cell models, the team analyzed approximately 1.6 million compounds using AI-based simulations and selected 5 optimal new drug candidates by considering blood-brain barrier permeability, toxicity, and gene binding affinity.

 

Among these, 2 candidates demonstrated approximately 100 times higher efficacy than the existing treatment phenytoin, paving the way for personalized treatment.

 

This study can be evaluated as a case of precision medicine that provides a practical alternative for intractable patients whose symptoms have not improved with existing treatments by combining the genetic characteristics of patients with rare genetic diseases with cell-based models and AI-based new drug discovery techniques.

 

Professor Kang explained, “This study is a case where patient-derived cell-based personalized drug discovery technology for patients with intractable epilepsy demonstrated an effect.

 

We plan to continue expanding our research to develop personalized precision therapies for patients with various genetic mutations, including SCN2A.” Professor Na emphasized, “This study demonstrates the practical applicability of precision medicine technology in the field of rare diseases.

 

We believe that the findings from this study will help establish an innovative treatment strategy that can be applied to patients with various genetic disorders in the future.” This study was conducted with support from the Public Health Technology Research Project promoted by the Ministry of Health and Welfare and the Korea Health Industry Development Institute, and was published in the 2025 issue of Computers in Biology and Medicine, a world-renowned academic journal in the field of medical information.