Finding the Cause of Infantile Nystagmus Syndrome Using Artificial Intelligence Deep Learning

Gangnam Severance Professors Jinwoo Han and Junwon Lee Team
Analyze All 3 Billion Base Sequences... Detect Mutations

Jinwoo Han (left) and Junwon Lee, Professors of Ophthalmology at Gangnam Severance Hospital.

[Asia Economy Reporter Lee Gwan-joo] Domestic researchers have utilized artificial intelligence (AI) deep learning to analyze the causes of rare eye diseases.

The ophthalmology team led by Professors Han Jin-woo and Lee Jun-won at Yonsei University Gangnam Severance Hospital announced on the 31st that they significantly improved the existing genetic testing method for analyzing the cause of infantile nystagmus syndrome, a rare eye disease, by integrating AI deep learning.

Infantile nystagmus syndrome is a rare eye disease characterized by involuntary eye movements horizontally, vertically, or in a mixed pattern in infants before six months of age. It is a disease with unclear causes, occurring in approximately one in every 2,000 people.

Recently, next-generation sequencing (NGS), a method for analyzing patients' genes, has become established, aiding in identifying causes, diagnosis, and treatment of infantile nystagmus syndrome. However, even with NGS techniques, 40-50% of patients still cannot have their causative mutations identified.

The research team introduced a method that uses deep learning with NGS to examine the entire genome. While previous methods were limited due to the vast scope, time, and cost of analyzing the whole genome, the team’s whole-genome analysis method inspects a broad region of approximately 3 billion base pairs. It can analyze non-transcribed regions that do not produce proteins and easily detect structural variations and regulatory region mutations in the genome.

The team conducted whole-genome analysis on four groups of infantile nystagmus syndrome patients and their families whose causes were not identified by previous methods, checking for mutations in the infantile nystagmus syndrome-associated gene 'FRMD7.' Through analysis, they discovered variants in the non-transcribed region of FRMD7 and confirmed that these variants caused errors in splicing (the process where introns are removed and exons are joined during the conversion of DNA to messenger RNA). Splicing errors are known to cause various diseases, including cancer and rare disorders.

Professor Han stated, “It is not easy to prove the association between mutations in non-transcribed regions and diseases, but AI deep learning enabled the prediction of splicing errors,” adding, “This study is expected to lay the foundation for utilizing AI deep learning and whole-genome analysis in patients with rare diseases whose causative mutations have not been found.”

This research was published in the international academic journal 'Translational Vision Science & Technology.'

This content was produced with the assistance of AI translation services.