Yuhui Hong
Postdoctoral Scholar at Noble Lab (2025 - Present)
University of Washington
Advised by Professor William Stafford Noble
Ph.D. in Computer Science (2020-2025)
Indiana University Bloomington
Advised by Professor Haixu Tang
I'm a postdoctoral researcher at the University of Washington working in computational mass spectrometry. Molecules leave signals in a mass spectrometer, such as fragmentation spectra, retention times, and collisional cross sections. Reading those signals back into a structure is hard: the mapping is ambiguous, and labeled data is scarce. I build machine learning models that identify molecules no reference library has seen before.
Please feel free to get in touch!
News
- 11/12/2026 TIMPANI has been accepted to the KDD 2027 AI for Sciences Track.
- 03/09/2026 Our work, De novo sequencing of chimeric spectra using Casanovo, has been selected for an oral presentation at ASMS 2026, San Diego.
- 02/11/2026 Invited to give a seminar talk at San Diego State University.
- 12/12/2025 FIDDLE is published in Nature Communications.
- 11/11/2025 Koina is published in Nature Communications, with 3DMolMS available as one of its models.
- 09/19/2025 Awarded the UW Data Science Fellowship at eScience Institute, University of Washington.
- 07/07/2025 I will join Noble Lab at University of Washington as a Postdoctoral Scholar in August 2025.
- 05/09/2025 Recipient of the Luddy Outstanding Research Award.
- 03/04/2025 Our work, A Task-Specific Transfer Learning Approach to Enhancing Small Molecule Retention Time Prediction with Limited Data, has been selected for an oral presentation at ASMS 2025, Baltimore.
Research interests
I develop machine learning methods for peptide and small-molecule identification from mass spectrometry data. My research asks what experimental measurements can tell us about molecules and how confidently we can interpret them.
-
Inferring peptide sequences and molecular formulasMy current work focuses on de novo peptide sequencing from chimeric tandem mass spectra, where fragments from multiple peptides appear in a single spectrum. I also develop methods to infer small-molecule formulas from fragmentation spectra.FIDDLE (Nature Communications, 2025) · De novo peptide sequencing (in progress, Noble Lab)
-
Predicting experimental measurements from molecular structureI build models that predict fragmentation spectra, retention times, and chromatographic enantioseparation from molecular structure. These predictions can expand spectral libraries and provide additional evidence for small-molecule identification.3DMolMS (Bioinformatics, 2023; Nature Communications, 2025) · 3DMolCSP (Analytical Chemistry, 2024) · TIMPANI (accepted by KDD 2027)
-
Assessing confidence in model predictionsI study how to estimate false discovery rates for de novo peptide sequencing when reference identifications are incomplete. I also investigate how confounding affects microbiome-based predictions and their interpretation.False discovery rate control for de novo identification (in progress, Noble Lab) · MicroKPNN-MT(Bioinformatics Advances, 2024) · MicroKPNN-CF(bioRxiv, 2025)
Selected publications
A few representative papers. For the complete list, please see my Google Scholar profile.
FIDDLE: a deep learning method for chemical formulas prediction from tandem mass spectra
Nature Communications, 16(1), 11102.
Machine learning in small-molecule mass spectrometry
Annual Review of Analytical Chemistry, 18.
Koina: democratizing machine learning for proteomics research
Nature Communications, 16(1), 9933.
Enhanced structure-based prediction of chiral stationary phases for chromatographic enantioseparation from 3D molecular conformations
Analytical Chemistry, 96(6), 2351–2359.