Ranking Metabolite Sets by Their Activity Levels

Related metabolites can be grouped into sets in many ways, e.g., by their participation series of chemical reactions (forming metabolic pathways), or based on fragmentation spectral similarities shared substructures. Understanding how such metabolite change relation to experimental factors incredibly useful the interpretation and understanding complex metabolomics data sets. However, available tools that are used perform this analysis not entirely suitable for untargeted measurements. Here, we present PALS (Pathway Activity Level Scoring), a Python library, command line tool, Web application performs ranking significantly changing over different conditions. The main algorithm is pathway level gene expression (PLAGE) factorisation method denoted as mPLAGE (PLAGE metabolomics). As an example application, analyse pathways tandem mass spectrometry patterns. A comparison with two other commonly methods (overrepresentation (ORA) set enrichment (GSEA)) also given reveals more robust missing features noisy than alternatives. further examples, applied human African trypanosomiasis, Rhamnaceae, American Gut Project data. In addition, normalisation have significant impact results, offers framework investigate this. freely from our project site.