paper.md
In adidevara/GOFIG: A Tool for Gene Ontology Enrichment Analysis and Visualization

title: 'GOFIG: A Tool for Gene Ontology Enrichment Analysis and Visualization' tags: - R - Gene Ontology - Enrichment Analysis - Data visualization authors: - name: Aditya K. Devarakonda orcid: 0000-0002-7014-3029 affiliation: "1" - name: Eric G. Rafalovsky affiliation: "2" - name: Tae-Jin Lee Ph.D. affiliation: "1" - name: Ashok Sharma Ph.D. orcid: 0000-0001-9597-4374 affiliation: "1" affiliations: - name: Center for Biotechnology and Genomic Medicine, Augusta University, Augusta, Georgia, United States index: 1 - name: Georgia Institute of Technology, Atlanta, Georgia, United States index: 2 date: 22 December 2020 bibliography: paper.bib

Summary

The advent of high throughput sequencing has fundamentally changed both basic science research and clinical medicine. RNA sequencing can be used to identify genetic changes and epigenetic regulation in a variety of cell types and conditions. A key tool in any cellular biologists arsenal to parse out these effects is RNA-seq differential expression analysis. While this is useful in evaluating the changes in cellular function, the output itself lists containing hundreds of gene identifiers and can be esoteric. A contextualized understanding of these results can be attained through a process known as Gene Ontology enrichment analysis. This process hierarchically categorizes genes into ontologies with terms that are much easier to understand and display.

Statement of need

GOFIG is an R package which allows for Gene Ontology enrichment analysis along with downstream ontology comparison and data visualization. While there are a existing programs which allow for Gene Ontology analysis, few are as easy to use while providing as much functionality as GOFIG.

The pipeline begins with a simple .csv file containing a list of gene identifications of almost any common type and fold change values. Using the org.Hs.eg.db package [@Carlson2020] in combination with the limma package [@Ritchie2015], the package is able to produce an enrichment analysis along with data regarding the number of up regulated and down regulated genes in each category. This feature only exists in a few packages for enrichment analysis currently. The output is saved as a .csv, and users can easily view the output while also marking ontologies which may be very similar to others and may not warrant display in the visuals.

For downstream visualization, a combination of the dplyr package [@Wickham2020] and ggplot2 [@Wickham2016] were used. There are two different types of bar plots which can be produced: a single bar plot showing the number of genes in the most significant ontologies along with points to represent the respective p-value of these ontologies and a back-to-back bar plot which provides the same information along with the subdivisions of up regulated and down regulated genes in each ontology. Another visualization which can be produced is a bubble plot that is depicted based on a normalization to display whether the ontology contains more up regulated or down regulated genes against the log of the p-values. The size of the bubbles is a representation of the number of genes in the ontology specified.

Aside from ease of use, GOFIG offers the unique functionality of comparing the overlapping ontologies between two different enrichment analyses and visualizing them through a Venn Diagram. This is done through the use of upstream enrichment analysis outputs along with VennDiagram package [@Chen2018].

The functionality along with the ease of use makes GOFIG and optimal tool for investigators looking to create data visuals to better represent the meaning of differential expression analysis data from RNA sequencing.