Computational biology / Machine learning / Web development

Shunichi Ito

M.S. student at Graduate School of Pharmaceutical Sciences, Kyoto University.
Passionate about revolutionizing drug discovery through life science × data science.

Projects

Profile

Shunichi Ito

M.S. student in Pharmaceutical Sciences at Kyoto University. Affiliated with the Biomolecular Measurement Laboratory, developing novel LC/MS data analysis methods using statistical models.

Research Interests: Chemoinformatics · AI for Drug Discovery · Diffusion Models
Expertise: EM Algorithm · Optimal Transport · Protein & Molecular LMs · Proteomics
Stack: Python ★★★ / TypeScript ★ / Go ★ / Ruby on Rails ★
Achievement: Kaggle Silver Medal · 20th Place (CIBMTR 2025) · Japan Statistical Society Certificate Grade 2
Language: Japanese / English
Hobbies: Futsal · Travel

Publications

Publications & Outputs

Poster figure: Multimodal drug discovery model architecture fusing ESM2 and ChemBERTa via cross-attention

40th Annual Conf. of JSAI, June 2026 · First author

Assessing Out-of-Distribution Generalization of Multimodal Foundation Models for Drug Discovery on Novel Protein Families

Abstract figure: Computational design of cardiomyocyte differentiation-inducing compounds from hiPSCs

Journal of Chemical Information and Modeling, 2024 · Co-author

Molecular Design for Cardiac Cell Differentiation Using a Small Data Set and Decorated Shape Features

Key visual: Gene network extraction with LLMs

Preferred Networks Tech Blog, 2025 · Summer Internship Output

Gene Network Extraction with LLMs: Pairwise Causal Inference and Prompt Design Experiments

Selected Work

Key Projects

Research 2024 - Present

LC/MS Data Analysis via Mixture-of-Mixtures Model Leveraging Sample Hierarchy

Developing an analysis tool that models complex LC/MS signal data as mixture distributions to identify and quantify proteins and peptides. Proposed a Mixture-of-Mixtures model that exploits sample hierarchy, and implemented an optimization algorithm using optimal transport as the objective function.

Internship Chugai Pharmaceutical

Assessing OOD Generalization of Multimodal Foundation Models on Novel Proteins

Evaluated the practical utility of multimodal foundation models for drug discovery tasks. Rigorously assessed out-of-distribution generalization to novel proteins from a practical standpoint, and proposed a new architecture that improves performance.

View presentation Read the article

Internship PFN

Gene Network Extraction with LLMs

Targeted 19 genes related to the breast cancer KEGG Pathway, performing pairwise directed-edge estimation with an LLM and integrating the results into a DAG. Improved F1 through iterative prompt design, revealing the potential to reconstruct comprehensive networks leveraging literature knowledge.

Read the article

Competition Kaggle

CIBMTR: Post-HCT Survival Prediction

Competed as a lab team in a post-hematopoietic stem cell transplantation prognosis prediction competition. Responsible for LightGBM tuning, earning a Silver Medal (20th place). Leveraged Kaplan-Meier transformation, a pairwise loss function, and seed averaging.

Competition page

Research iCeMS / KyotoU

Computational Design of Cardiomyocyte Differentiation-Inducing Compounds from hiPSCs

Participated as a part-time researcher at iCeMS, Kyoto University. Predicted differentiation-inducing compounds from limited data using ML. Responsible for fragment decomposition, novel compound generation, and PDB file creation. Experimental validation confirmed cardiomyocyte differentiation; published in a peer-reviewed international journal.

Read the paper

Productivity LINE Bot

Lab Equipment Maintenance Notification Bot

Built a LINE Bot using Google Apps Script, Google Calendar API, and LINE API that notifies assigned members the day before their LC/MS/MS maintenance shift. Actively used by lab members on a daily basis.

Experience

Research & Development Experience

Apr 2024 – Present

M.S. Course, Graduate School of Pharmaceutical Sciences, Kyoto University

Affiliated with the Biomolecular Measurement Laboratory, working on developing analysis methods for proteomics LC/MS data using mixture distribution models.

May 2026 – Present

RUTILEA Inc. — Part-time

Working on the development of LLM applications.

Aug – Oct 2025

Chugai Pharmaceutical — Internship

Evaluated drug discovery tasks using protein language models, molecular language models, and multimodal foundation models. Analyzed practical limitations and developed a new architecture to improve predictive accuracy.

Aug – Sep 2025

Preferred Networks — Summer Internship

Worked on gene network extraction with LLMs, designing and evaluating a method to infer causal relationships from literature knowledge.

Feb – Mar 2025

Advemto Ltd. — Internship

Built a machine learning model to predict fluorescent labels from 2D tensor fluorescence spectrum data. Proposed and implemented a preprocessing approach that preserves 2D data structure.

Oct 2022 – Mar 2025

DMG Mori / WALC — Long-term Internship

Worked on medical image segmentation, health checkup data analysis, RAG chatbot development, defect detection with SAM, schedule optimization with OR-Tools, and safety behavior monitoring with YOLO.

Sep 2023 – Mar 2024

Strategy and Partner — Part-time Engineer

Contributed to a web application built with the OpenAI API. Implemented backend APIs in Go and improved the chat UI in Next.js, including Markdown rendering and streaming response delivery.

Skills

Expertise

✔ Broad knowledge and experience across statistics, machine learning, deep learning, and computer science.
✔ Wide domain knowledge in pharmaceutical and life sciences, with expertise in proteomics.

Programming

Python, TypeScript, Go, Git, Docker, Linux

Machine Learning

PyTorch, Scikit-Learn, XGBoost, LightGBM, Transformers, YOLO, RAG

Domain

Computational drug discovery, Proteomics, LC/MS/MS, Molecular / Protein language models

Web

React, Next.js, Gin, Ruby on Rails, OpenAI API