Bioinformatics Analysis of Protein Secretion in Plants

In silico prediction of the secretome from the invasive neurotoxic marine dinoflagellate Alexandrium catenella

Alexandrium catenella, a marine dinoflagellate responsible for harmful algal blooms (HABs), proliferates with greater frequency, distribution and intensity, in disturbed marine coastal ecosystems. The proteins secreted into seawater may play a crucial role in maintaining this dinoflagellate in these ecosystems, but this possibility has never been investigated before. In this study, the A. catenella secretome was predicted from its transcriptome by combining several bioinformatics tools. Our results predict a secretome of 2 779 proteins, among which 79% contain less than 500 amino acids, suggesting that most secreted proteins are short in length. The predicted secretome includes 963 proteins (35%) with Pfam domains: 773 proteins with one Pfam domain and 190 proteins with two or more Pfam domains. Their functional annotation showed that they are mainly involved in (i) proteolysis, (ii) stress responses and (iii) primary metabolism. In addition, 47% of the secreted proteins appear to be enzymes, primarily peptidases, known to be biologically active in the extracellular medium during stress responses. Finally, this study provides a wealth of candidates of proteins secreted by A. catenella, which may interact with the marine environment and help this dinoflagellate develop in various environmental conditions.

MetaGOmics: A Web-Based Tool for Peptide-Centric Functional and Taxonomic Analysis of Metaproteomics Data

Metaproteomics is the characterization of all proteins being expressed by a community of organisms in a complex biological sample at a single point in time. Applications of metaproteomics range from the comparative analysis of environmental samples (such as ocean water and soil) to microbiome data from multicellular organisms (such as the human gut). Metaproteomics research is often focused on the quantitative functional makeup of the metaproteome and which organisms are making those proteins. That is: What are the functions of the currently expressed proteins? How much of the metaproteome is associated with those functions? And, which microorganisms are expressing the proteins that perform those functions? However, traditional protein-centric functional analysis is greatly complicated by the large size, redundancy, and lack of biological annotations for the protein sequences in the database used to search the data. To help address these issues, we have developed an algorithm and web application (dubbed "MetaGOmics") that automates the quantitative functional (using Gene Ontology) and taxonomic analysis of metaproteomics data and subsequent visualization of the results. MetaGOmics is designed to overcome the shortcomings of traditional proteomics analysis when used with metaproteomics data. It is easy to use, requires minimal input, and fully automates most steps of the analysis-including comparing the functional makeup between samples. MetaGOmics is freely available at https://www.yeastrc.org/metagomics/.