Mass spectral similarity for untargeted metabolomics data analysis of complex mixtures

While in nucleotide sequencing, the analysis of DNA from complex mixtures of organisms is common, this is not yet true for mass spectrometric data analysis of complex mixtures. The comparative analyses of mass spectrometry data of microbial communities at the molecular level is difficult to perform, especially in the context of a host. The challenge does not lie in generating the mass spectrometry data, rather much of the difficulty falls in the realm of how to derive relevant information from this data. The informatics based techniques to visualize and organize datasets are well established for metagenome sequencing; however, due to the scarcity of informatics strategies in mass spectrometry, it is currently difficult to cross correlate two very different mass spectrometry data sets from microbial communities and their hosts. We highlight that molecular networking can be used as an organizational tool of tandem mass spectrometry data, automated database search for rapid identification of metabolites, and as a workflow to manage and compare mass spectrometry data from complex mixtures of organisms. To demonstrate this platform, we show data analysis from hard corals and a human lung associated with cystic fibrosis.