| Timetable |
| Download Conference Program |
| Download All Abstracts |
| Zoom Access |
| Corporate Program |
Poster Presentations
Day 3, June 24（Tue.）
Room P (Maesato East, Foyer, Ocean Wing)
- 3P-AM-31
Shin (Neo)-MassBank Project: Enriching MassBank records using human metabolome datasets with FDR-controlled metabolite annotation
(1Osaka Univ., 2Keio Univ., 3Kyushu Univ., 4Tokyo Univ. Agr. Tech., 5Niigata Univ.)
oFumio Matsuda1, Ryosuke Hayasaka2, Taihei Torigoe3, Yushi Takahashi5, Takaki Oka4, Yuki Matsuzawa4, Kozo Nishida4, Masatomo Takahashi3, Akiyasu Yoshizawa5, Takato Kiuchi5, Hiroshi Tsugawa4, Akiyoshi Hirayama2, Yoshihiro Izumi3, Shujiro Okuda5
The Shin (Neo)-MassBank project aims to enhance MassBank records by generating high-quality MS2 spectra from human metabolome DDA datasets. The project introduces three key resources: a human sample-derived MS2 spectra library, a method for controlling false discovery rates (FDR), and comprehensive MassBank record support in MS-DIAL software.
To build the MS2 spectra library, human metabolome datasets were obtained from Metabolomics Workbench and MetaboLights. Using in-house Python3 scripts for spectrum averaging and annotation tools like SIRIUS, CFM-ID, MetFrag, and MS-FINDER, the team produced 2,873 high-precision spectra from 365,333 raw MS2 spectra. Among these, 73 spectra achieved confidence level 2a and 293 spectra reached level 3. These annotated spectra are available through the MassBank Human repository.
To improve FDR control in metabolite identification, the project evaluated four methods, with the second-rank method proving most effective in achieving an FDR close to 0.05.
Additionally, MS-DIAL now fully supports the MassBank record format, allowing seamless data export and improved visualization. These advancements collectively enhance data quality, annotation accuracy, and usability in metabolomics research.