Auto-GO: Reproducible, Robust and High Quality Ontology Enrichment Visualizations

Bioinformatics and Data Science are routinely challenged to take out intelligible results from huge amounts of data. These results, in turn, are conveyed through plots and visualizations that should be easily reproducible for scientific soundness and ethical reasons. This process is of critical importance in Genomics, when dealing with large multi-omics datasets. One of the final steps of these analyses is Gene Set enrichment, where web tools represent a valuable resource but not a reliable surrogate for standardized, high-quality visualizations. Here we present Auto-GO, a software framework that proposes a standardization of the Gene Functional Enrichment process along with an R package able to produce high-quality visualization in an automated manner, improving the reproducibility of the whole analytical process. We present three use cases in Cancer Transcriptomics and Epigenomics datasets as a proof-of-concept to visualize Multiple Differential Expression and Single Sample Gene Set Enrichment Analysis. Finally, we show how it can be employed to shear light on publicly available datasets, even in small casuistry of Rare Cancers.