MassSpecGym

MassSpecGym provides a dataset and benchmark for the discovery and identification of new molecules from tandem mass spectrometry (MS/MS) spectra. The provided challenges abstract the process of scientific discovery of new molecules from biological and environmental samples into well-defined machine learning problems. P

Papers using MassSpecGym (15)

MassSpecGym: A benchmark for the discovery and identification of molecules2024 · 50 citesJESTR: Joint Embedding Space Technique for Ranking Candidate Molecules for the Annotation of Untargeted Metabolomics Data2024 · 13 citesMADGEN: Mass-Spec attends to De Novo Molecular generation2025 · 3 citesTest-Time Tuned Language Models Enable End-to-end De Novo Molecular Structure Generation from MS/MS Spectra2025 · 2 citesDe Novo Molecular Generation from Mass Spectra via Many-Body Enhanced Diffusion2026 · 1 citesOne Small Step with Fingerprints, One Giant Leap for De Novo Molecule Generation from Mass Spectra2025 · 2 citesUnlocking High-Fidelity Molecular Generation from Mass Spectra via Dual-Stream Line Graph Diffusion2026MSAlign: Aligning Molecule and Mass Spectra Foundation Models for Metabolite Identification2026FRIGID: Scaling Diffusion-Based Molecular Generation from Mass Spectra at Training and Inference Time2026When should we trust the annotation? Selective prediction for molecular structure retrieval from mass spectra2026How well can off-the-shelf LLMs elucidate molecular structures from mass spectra using chain-of-thought reasoning?2026SpecBridge: Bridging Mass Spectrometry and Molecular Representations via Cross-Modal Alignment2026Breaking the Modality Barrier: Generative Modeling for Accurate Molecule Retrieval from Mass Spectra2025General Intelligence-based Fragmentation (GIF): A framework for peak-labeled spectra simulation2025MS-BART: Unified Modeling of Mass Spectra and Molecules for Structure Elucidation2025