2023 Fiscal Year Final Research Report

Development of deep learning methodology in RNA-seq analysis and application to sarcoma treatment

Research Project

PDF

Project/Area Number	20K09453
Research Category	Grant-in-Aid for Scientific Research (C)
Allocation Type	Multi-year Fund
Section	一般
Review Section	Basic Section 56020:Orthopedics-related
Research Institution	The University of Tokyo
Principal Investigator	Kobayashi Hiroshi 東京大学, 医学部附属病院, 講師 (20407951)
Co-Investigator(Kenkyū-buntansha)	篠田裕介埼玉医科大学, 医学部, 教授 (80456110) 谷口優樹東京大学, 医学部附属病院, 講師 (80722165)
Project Period (FY)	2020-04-01 – 2024-03-31
Keywords	骨軟部腫瘍 / 未分化多形肉腫 / 深層学習
Outline of Final Research Achievements	Sarcoma is a malignant tumor that originates from mesenchymal cells such as muscle and fat, it can be difficult to diagnose because it is rare and consists of about 70 histological types. Undifferentiated pleomorphic sarcoma (UPS) is the most common sarcoma that occurs in soft tissue, and is diagnosed after excluding other ``tumors with a clear differentiation direction or specific fusion genes''. This study aimed to reclassify, stratify, and elucidate the pathology of UPS using deep learning analyzing gene expression of UPS. At present, we have not identified any characteristics that can distinguish it from other sarcomas, but we will further investigate clinical information and stratify UPS to identify abnormalities that may lead to clinical applications and new treatments.
Free Research Field	骨軟部腫瘍
Academic Significance and Societal Importance of the Research Achievements	未分化多形肉腫は肉腫の中でも高頻度に発生する高悪性度の肉腫である。未分化多形肉腫といっても臨床的には治療への反応、予後が異なるものが含まれていると考えられる。そのため、遺伝子発現データを深層学習を用いて未分化多形肉腫を再分類し、適切な治療方針、新規治療法の開発を目指した新規性のある研究である。本研究によって、臨床情報の特徴を抽出する未分化多形肉腫の層別化には完全には至らなかったが、引き続きデータを集積して臨床応用、新規治療開発に繋げえられるようにさらなる解析を行なっていく予定である。