GENE HUNTER: A NOVEL WEB-TOOL TO VISUALIZE SIMULTANEOUSLY DIFFERENTIAL GENES EXPRESSION ACROSS MULTIPLE DATASETS

JUAN BIZZOTTO; ELBA VAZQUEZ; AGUSTINA SABATER; JAVIER COTIGNOLA; ROSARIO LAVIGNOLLE; PABLO SANCHIS; GERALDINE GUERON

Virtual

Congreso; Reunión Anual de la Sociedad Argentina de Investigación Clínica 2021; 2021

SAIC

Bioinformatics are becoming a prominent component of cancer research, resulting in an increased number of publicly available datasets to study and making it possible to evaluate gene expression across a variety of populations or disease stages. However, having to search for a specific gene in multiple datasets can often be time consuming and unfriendly to novel bioinformaticians. Our aim was to develop a user-friendly search tool to access simultaneously data of differential expression analyses for a particular gene or even gene signatures across multiple datasets.For this purpose, 7 publicly available transcriptomic prostate cancer (PCa) datasets (n total= 875) were selected to either be used by researchers interested in PCa or as an example to understand the tool before adapting it to their specific need, and differential expression analyses were performed in R, using the limma package. A Shiny-based tool, that can be accessed through a user interface, was then built to execute the search.Our Shiny app includes a search bar that allows researchers to look for either a specific gene or a family of genes within all the datasets. Search results are presented in tables containing information on the comparison made for the analysed dataset, the gene ID and symbol, t and B statistical parameters, the log Fold Change, the p-value, adjusted p-value and the dataset?s GSE identifier. Additionally, users can also plot different variables to visualize gene expression in all selected datasets more easily. Among the options, clinical significance of a gene can be assessed by overall survival and Kaplan-Meier plots.In summary, Gene Hunter is a novel Shiny-based tool with the potential to ease high-throughput analyses in basic cancer research. It does so by providing the opportunity to explore differential expression between tumoral conditions, while straddling the limits of individual studies. This platform has the potential to extend to dataset comprehending other pathologies.