Extracting pathway-level insights from untargeted metabolomics remains a major challenge due to data complexity and lack of targeted reference points. This talk presents a computational toolkit designed to enhance pathway inference from untargeted metabolomics data, enabling more precise biological interpretation.