京都大学 京都大学 大学院医学研究科 人間健康科学系専攻 ビッグデータ医科学分野

SEUNGEON LEE

PORTFOLIO

Researching AI-driven drug discovery for medium-sized peptides, focusing on peptide language models and property prediction.

ABOUT

Seungeon Lee

LEE SEUNGEON  /  李 昇彦
Graduate School of Medicine, Kyoto University
Big Data Medical Science, Okuno Laboratory

AI Drug Discovery / Cheminformatics / Medium-sized Peptides / Machine Learning

I am a Master's student at the Graduate School of Medicine, Kyoto University, working in the Okuno Laboratory on AI-driven drug discovery. My current research focuses on transformer-based language models for medium-sized peptide property prediction, bridging the gap between small-molecule and protein informatics.

SELECTED PUBLICATIONS

Journal of Chemical Information and Modeling (JCIM) 2026
HELM-BERT: Topology-aware representations for chemically modified peptides
Seungeon Lee, Takuto Koyama, Itsuki Maeda, Shigeyuki Matsumoto, Yasushi Okuno
DOI
Journal of Cheminformatics 2026
Empowering federated learning for robust compound–protein interaction prediction across heterogeneous cross-pharma domains
Takuto Koyama, Hiroaki Iwata, Ryosuke Kojima, Takao Otsuka, Aki Hasegawa, Seungeon Lee, Hiroshi Ueda, Toshiharu Morimoto, Ryoko Sasaki, Nao Torimoto, Sei Murakami, Manabu Tojo, Teruki Honma, Shigeyuki Matsumoto, Yasushi Okuno
DOI PubMed PMC
Download BibTeX

BIOGRAPHY

2025.04 – Present

Graduate Student (M.S.)

Kyoto University, Graduate School of Medicine, Division of Human Health Sciences (Okuno Laboratory)

2025.03

Undergraduate (B.Sc.)

Kwansei Gakuin University, School of Science and Technology, Department of Bioscience (Miura Laboratory)

BIBLIOGRAPHY

ARTICLES

  1. HELM-BERT: Topology-aware representations for chemically modified peptides Seungeon Lee, Takuto Koyama, Itsuki Maeda, Shigeyuki Matsumoto, Yasushi Okuno. (2026). Journal of Chemical Information and Modeling (JCIM) IN PRESS
  2. Empowering federated learning for robust compound–protein interaction prediction across heterogeneous cross-pharma domains Takuto Koyama, Hiroaki Iwata, Ryosuke Kojima, Takao Otsuka, Aki Hasegawa, Seungeon Lee, Hiroshi Ueda, Toshiharu Morimoto, Ryoko Sasaki, Nao Torimoto, Sei Murakami, Manabu Tojo, Teruki Honma, Shigeyuki Matsumoto, Yasushi Okuno. (2026). Journal of Cheminformatics

PREPRINTS

  1. HELM-BERT: A Transformer for Medium-sized Peptide Property Prediction Seungeon Lee, Takuto Koyama, Itsuki Maeda, Shigeyuki Matsumoto, Yasushi Okuno. (2025). arXiv PREPRINT

PRESENTATIONS

  1. ゼロからわかるChatGPT — Practical AI Adoption for Healthcare Teams Yokawa Hospital (吉川病院), Miki, Hyogo (Invited). (2026.03.09) INVITED ゼロからわかるChatGPT — Practical AI Adoption for Healthcare Teams
  2. HELM-BERT: A Transformer for Medium-sized Peptide Property Prediction CBI Society Annual Meeting 2025 (Oral, O03-02). (2025.10.29) ORAL HELM-BERT: A Transformer for Medium-sized Peptide Property Prediction

AWARDS

WRITINGS

RECENT WORKS

2026.05.07 WRITING 低分子創薬AIにおける表現学習・相互作用理解・候補分子設計の進展 — 月刊バイオインダストリー（CMC出版）2026年7月号 特集「AI創薬の最前線」 IN PRESS 2026.05.07 ARTICLE HELM-BERT: Topology-aware representations for chemically modified peptides — Journal of Chemical Information and Modeling (JCIM) IN PRESS 2026.03.09 TALK ゼロからわかるChatGPT — Practical AI Adoption for Healthcare Teams — Yokawa Hospital (吉川病院), Miki, Hyogo (Invited) INVITED 2026.01.10 ARTICLE Empowering federated learning for robust compound–protein interaction prediction across heterogeneous cross-pharma domains — Journal of Cheminformatics 2025.12.29 PREPRINT HELM-BERT: A Transformer for Medium-sized Peptide Property Prediction — arXiv 2025.10.30 AWARD Excellent Presentation Award — CBI Society Annual Meeting 2025 2025.10.29 TALK HELM-BERT: A Transformer for Medium-sized Peptide Property Prediction — CBI Society Annual Meeting 2025 (Oral, O03-02) ORAL

SKILL

code

Python

Research experience with Python programming
Deep learning implementation with PyTorch

psychology

Machine Learning

Transformer-based language models
Peptide property prediction & federated learning

biotech

Cheminformatics

HELM notation for peptide representation
Compound-protein interaction modeling

language

Korean

Native speaker (mother tongue)

language

English

TOEIC 880

language

Japanese

JLPT N1

CONTACT

Feel free to reach out via SNS or email.

ORCID Google Scholar ResearchGate GitHub LinkedIn @flansma mail lee.seungeon.62y[at]st.kyoto-u.ac.jp
expand_less