Use of Molecular Networking for Compound Annotation in Metabolomics

The metabolomics community and natural product researchers usually use hyphenated analytical techniques to investigate the metabolic profile in complex matrices. Modern analytical instruments used in metabolomic strategies usually generally provide a large amount of data, and these results processing can be optimized by computational tools. The use of classical metabolomics tools, such as consulting the reference databases represents one of the commonly used strategies, however, this approach can lead to a considerable number of unidentified metabolites, in addition to being a time-consuming task if performed manually. In 2012, an important computational strategy for metabolomic study was developed at The Dorrestein Laboratory by Pieter Dorrestein's research group, this approach called molecular networking is generated by the GNPS platform (Global Natural Products Social) based on the hypothesis that structurally related molecules produce similar fragmentation patterns, thus they are integrated in the same group. The analysis of molecular networks makes possible to annotate known molecules based on the identification of potential similarities between all mass spectra within the data set, and to extend annotations to unknown molecules, but with very similar fragmentation patterns. The use of advanced molecular network annotation tools such as the MolNetEnhancer, the network annotation propagation (NAP) and the MS2LDA-MOTIF, contribute to the identification of unannotated ions, being of great value for metabolomic studies. These tools allow the simultaneous analysis of numerous mass spectra obtained in several experiments, reducing the data processing time. This work aims to present classical and advanced computational tools available for research in metabolomics, as well as to discuss the analysis of necessary parameters for this molecular profile approach.