Search for: All records

Recent developments in molecular networking have expanded our ability to characterize the metabolome of diverse samples that contain a significant proportion of ion features with no mass spectral match to known compounds. Manual and tool-assisted natural annotation propagation is readily used to classify molecular networks; however, currently no annotation propagation tools leverage consensus confidence strategies enabled by hierarchical chemical ontologies or enable the use of new in silico tools without significant modification. Herein we present ConCISE (Consensus Classifications of In Silico Elucidations) which is the first tool to fuse molecular networking, spectral library matching and in silico class predictions to establish accurate putative classifications for entire subnetworks. By limiting annotation propagation to only structural classes which are identical for the majority of ion features within a subnetwork, ConCISE maintains a true positive rate greater than 95% across all levels of the ChemOnt hierarchical ontology used by the ClassyFire annotation software (superclass, class, subclass). The ConCISE framework expanded the proportion of reliable and consistent ion feature annotation up to 76%, allowing for improved assessment of the chemo-diversity of dissolved organic matter pools from three complex marine metabolomics datasets comprising dominant reef primary producers, five species of the diatom genus Pseudo-nitzchia, and stromatolite sediment samples. 

ABSTRACT Dissolved organic matter (DOM) comprises diverse compounds with variable bioavailability across aquatic ecosystems. The sources and quantities of DOM can influence microbial growth and community structure with effects on biogeochemical processes. To investigate the chemodiversity of labile DOM in tropical reef waters, we tracked microbial utilisation of over 3000 untargeted mass spectrometry ion features exuded from two coral and three algal species. Roughly half of these features clustered into over 500 biologically labile spectral subnetworks annotated to diverse structural superclasses, including benzenoids, lipids, organic acids, heterocyclics and phenylpropanoids, comprising on average one‐third of the ion richness and abundance within each chemical class. Distinct subsets of these labile compounds were exuded by algae and corals during the day and night, driving differential microbial growth and substrate utilisation. This study expands the chemical diversity of labile marine DOM with implications for carbon cycling in coastal environments. 

Abstract The dominant benthic primary producers in coral reef ecosystems are complex holobionts with diverse microbiomes and metabolomes. In this study, we characterize the tissue metabolomes and microbiomes of corals, macroalgae, and crustose coralline algae via an intensive, replicated synoptic survey of a single coral reef system (Waimea Bay, Oʻahu, Hawaii) and use these results to define associations between microbial taxa and metabolites specific to different hosts. Our results quantify and constrain the degree of host specificity of tissue metabolomes and microbiomes at both phylum and genus level. Both microbiome and metabolomes were distinct between calcifiers (corals and CCA) and erect macroalgae. Moreover, our multi-omics investigations highlight common lipid-based immune response pathways across host organisms. In addition, we observed strong covariation among several specific microbial taxa and metabolite classes, suggesting new metabolic roles of symbiosis to further explore.