BackgroundAssembly and function of neuronal synapses require the coordinated expression of a yet undetermined set of genes. Although roughly a thousand genes are expected to be important for this function in Drosophila melanogaster, just a few hundreds of them are known so far.ResultsIn this work we trained three learning algorithms to predict a “synaptic function” for genes of Drosophila using data from a whole-body developmental transcriptome published by others. Using statistical and biological criteria to analyze and combine the predictions, we obtained a gene catalogue that is highly enriched in genes of relevance for Drosophila synapse assembly and function but still not recognized as such.ConclusionsThe utility of our approach is that it reduces the number of genes to be tested through hypothesis-driven experimentation.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-015-1888-3) contains supplementary material, which is available to authorized users.

Supplementary data are available at Bioinformatics online.

BackgroundAssembly and function of neuronal synapses require the coordinated expression of a yet undetermined set of genes. Previously, we had trained an ensemble machine learning model to assign a probability of having synaptic function to every protein-coding gene in Drosophila melanogaster. This approach resulted in the publication of a catalogue of 893 genes which we postulated to be very enriched in genes with a still undocumented synaptic function. Since then, the scientific community has experimentally identified 79 new synaptic genes. Here we use these new empirical data to evaluate our original prediction. We also implement a series of changes to the training scheme of our model and using the new data we demonstrate that this improves its predictive power. Finally, we added the new synaptic genes to the training set and trained a new model, obtaining a new, enhanced catalogue of putative synaptic genes.ResultsThe retrospective analysis demonstrate that our original catalogue was significantly enriched in new synaptic genes. When the changes to the training scheme were implemented using the original training set we obtained even higher enrichment. Finally, applying the new training scheme with a training set including the 79 new synaptic genes, resulted in an enhanced catalogue of putative synaptic genes. Here we present this new catalogue and announce that a regularly updated version will be available online at: http://synapticgenes.bnd.edu.uyConclusionsWe show that training an ensemble of machine learning classifiers solely with the whole-body temporal transcription profiles of known synaptic genes resulted in a catalogue with a significant enrichment in undiscovered synaptic genes. Using new empirical data provided by the scientific community, we validated our original approach, improved our model an obtained an arguably more precise prediction. This approach reduces the number of genes to be tested through hypothesis-driven experimentation and will facilitate our understanding of neuronal function.Availability http://synapticgenes.bnd.edu.uy

The function of most genes is unknown. The best results in automated function prediction are obtained with machine learning-based methods that combine multiple data sources, typically sequence derived features, protein structure and interaction data. Even though there is ample evidence showing that a gene’s function is not independent of its location, the few available examples of gene function prediction based on gene location rely on sequence identity between genes of different organisms and are thus subjected to the limitations of the relationship between sequence and function. Here we predict thousands of gene functions in five model eukaryotes (Saccharomyces cerevisiae, Caenorhabditis elegans, Drosophila melanogaster, Mus musculus and Homo sapiens) using machine learning models exclusively trained with features derived from the location of genes in the genomes to which they belong. Our aim was not to obtain the best performing method to automated function prediction but to explore the extent to which a gene's location can predict its function in eukaryotes. We found that our models outperform BLAST when predicting terms from Biological Process and Cellular Component Ontologies, showing that, at least in some cases, gene location alone can be more useful than sequence to infer gene function.