MassGAT: a graph-based collective learning approach for untargeted detection and annotation of LC-MS data
Motivation: The untargeted processing of Liquid Chromatography coupled to High-Resolution Mass Spectrometry (LC-HRMS) data is a major challenge for the comprehensive and robust characterisation of metabolites. In particular, peak detection and annotation are two challenging tasks due to the size and complexity of the data, that are currently addressed independently in existing pipelines, without…
MassGAT: A Graph-Based Approach for Unbiased Detection and Annotation of LC-MS Data
The challenge of processing Liquid Chromatography coupled to High-Resolution Mass Spectrometry (LC-HRMS) data in an unbiased manner has been a major hurdle in the comprehensive characterization of metabolites. Current pipelines tackle detection and annotation separately, ignoring the redundancy between ion species of the same compound.
In response, researchers have developed an innovative and efficient approach called MassGAT. This model represents signals likely originating from the same molecule as a graph and uses a Graph Attention Network (GAT) to infer the validity of peaks and their connections within each component. The results demonstrate that MassGAT outperforms existing methods in both detection and annotation, as shown on real data sets.
The MassGAT open-source Python module is now publicly available, accessible at https://github.com/odisce/MassGAT.
Written by urgent.news from bioRxiv's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
