maxiM/Ze: An Image Recognition Approach for Visualizing and Processing Mass Spectrometry Based Metabolomics Data

Informatics is essential in metabolomics to analyze and interpret complex data for the advancement of biological insights. However, many current data-processing tools are time-consuming, require careful parameter selection, and depend heavily on user expertise, making reproducibility a challenge. To address these challenges, we developed maxiM/Ze, a Python-based application that utilizes image recognition algorithms to process liquid chromatography-high resolution mass spectrometry (LC-HRMS) metabolomics data prior to statistical analysis. The software implements an automated sequential pipeline that includes mass detection, extracted ion chromatogram (EIC) generation, peak alignment, and data visualization. By converting extracted ion chromatograms into PNG images, maxiM/Ze applies image processing techniques from OpenCV, including Canny edge detection, watershed segmentation, and Pearson correlation-based clustering, to align peaks across samples with minimal user input. Validation against Compound Discoverer 3.4 and mzmine 4.8.30 using eight replicate pooled plasma samples demonstrated competitive feature detection (12,067 features), annotation (219 unique compounds), and reproducibility (median CV of 35.8%) across platforms. The application is prepared for release on both Mac OS and Windows platforms, with the goal of improving reproducibility in metabolomics data analysis.