Search

Moonlighting proteins (MPs) are a special type of protein with multiple independent functions. MPs play vital roles in cellular regulation, diseases, and biological pathways. At present, very few MPs have been discovered by biological experiments. Due to the lack of data sample, computation-based methods to identify MPs are limited. Currently, there is no de-novo prediction method for MPs. Therefore, systematic research and identification of MPs are urgently required. In this paper, we propose a multimodal deep ensemble learning architecture, named MEL-MP, which is the first de novo computation model for predicting MPs. First, we extract four sequence-based features: primary protein sequence information, evolutionary information, physical and chemical properties, and secondary protein structure information. Second, we select specific classifiers for each kind of feature. Finally, we apply the stacked ensemble to integrate the output of each classifier. Through comprehensive model selection and cross-validation experiments, it is shown that specific classifiers for specific feature types can achieve superior performance. For validating the effectiveness of the fusion-based stacked ensemble, different feature fusion strategies including direct combination and a multimodal deep auto-encoder are used for comparative purposes. MEL-MP is shown to exhibit superior prediction performance (F-score = 0.891), surpassing the existing machine learning model, MPFit (F-score = 0.784). In addition, MEL-MP is leveraged to predict the potential MPs among all human proteins. Furthermore, the distribution of predicted MPs on different chromosomes, the evolution of MPs, the association of MPs with diseases, and the functional enrichment of MPs are also explored. Finally, for maximum convenience, a user-friendly web server is available at: http://ml.csbg-jlu.site/mel-mp/.

Gene duplication is a recurring phenomenon in genome evolution and a major driving force in the gain of biological functions. Here, we examine the role of gene duplication in the origin and maintenance of moonlighting proteins, with special focus on functional redundancy and innovation, molecular tradeoffs, and genetic robustness. An overview of specific examples--mainly from yeast--suggests a widespread conservation of moonlighting behavior in duplicate genes after long evolutionary times. Dosage amplification and incomplete subfunctionalization appear to be prevalent in the maintenance of multifunctionality. We discuss the role of gene-expression divergence and paralog responsiveness in moonlighting proteins with overlapping biochemical properties. Future studies analyzing multifunctional genes in a more systematic and comprehensive manner will not only enable a better understanding of how this emerging class of protein behavior originates and is maintained, but also provide new insights on the mechanisms of evolution by gene duplication.

The examples above illustrate just some of the ways in which the continuing study of moonlighting proteins is important: developing novel therapeutics, identifying previously unknown cellular processes, elucidating the connections among biochemical pathways, understanding novel protein mechanisms, improving the prediction of protein functions, and providing information about the evolution of protein structure and function as well as examples for the design of new proteins. The references listed above and especially the collection of papers in this Research Topic are recommended for providing more in-depth examples and analysis of these topics. Conflict of interest statement The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

International audience; Moonlighting proteins are defined by their involvement in multiple, unrelated functions. The computational prediction of such proteins requires a formal method of assessing the similarity of cellular processes, for example, by identifying dissimilar Gene Ontology terms. While many measures of Gene Ontology term similarity exist, most depend on abstract mathematical analyses of the structure of the GO tree and do not necessarily represent the underlying biology. Here, we propose two metrics of GO term functional dissimilarity derived from biological information, one based on the protein annotations and the other on the interactions between proteins. They have been collected in the PrOnto database, a novel tool which can be of particular use for the identification of moonlighting proteins. The database can be queried via an web-based interface which is freely available at http://tagc.univ-mrs.fr/pronto.