MAGNET: A web-based application for gene set enrichment analysis using macrophage data sets

PLOS ONE RESEARCH ARTICLE MAGNET: A web-based application for gene set enrichment analysis using macrophage data sets Shang-Yang Chen ID, Gaurav Gadhvi, Deborah R. Winter ID* Division of Rheumatology, Feinberg School of Medicine, Northwestern University, Chicago, IL, United States of America a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Chen S-Y, Gadhvi G, Winter DR (2023) MAGNET: A web-based application for gene set enrichment analysis using macrophage data sets. PLoS ONE 18(1): e0272166. https://doi.org/ 10.1371/journal.pone.0272166 Editor: Jishnu Das, University of Pittsburgh, UNITED STATES Received: November 4, 2021 Accepted: July 13, 2022 Published: January 11, 2023 Copyright: © 2023 Chen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: No new genomic data was generated as part of this study. Full analytical results are included in supplementary tables. Funding: SC is grateful to the American Heart Association for the predoctoral fellowship. DRW is supported by the Arthritis National Research Foundation (ANRF), the American Heart Association (AHA: 18CDA34110224), the Scleroderma Foundation, the American Lung Association (ALA), the American Thoracic Society (ATS), the American Federation for Aging (AFAR), * Abstract Characterization of gene lists obtained from high-throughput genomic experiments is an essential task to uncover the underlying biological insights. A common strategy is to perform enrichment analyses that utilize standardized biological annotations, such as GO and KEGG pathways, which attempt to encompass all domains of biology. However, this approach provides generalized, static results that may fail to capture subtleties associated with research questions within a specific domain. Thus, there is a need for an application that can provide precise, relevant results by leveraging the latest research. We have therefore developed an interactive web application, Macrophage Annotation of Gene Network Enrichment Tool (MAGNET), for performing enrichment analyses on gene sets that are specifically relevant to macrophages. Using the hypergeometric distribution, MAGNET assesses the significance of overlapping genes with annotations that were curated from published manuscripts and data repositories. We implemented numerous features that enhance utility and user-friendliness, such as the simultaneous testing of multiple gene sets, different visualization options, option to upload custom datasets, and downloadable outputs. Here, we use three example studies compared against our current database of ten publications on mouse macrophages to demonstrate that MAGNET provides relevant and unique results that complement conventional enrichment analysis tools. Although specific to macrophage datasets, we envision MAGNET will catalyze developments of similar applications in other domains of interest. MAGNET can be freely accessed at the URL https:// magnet-winterlab.herokuapp.com. Website implemented in Python and PostgreSQL, with all major browsers supported. The source code is available at https://github.com/ sychen9584/MAGNET. Introduction Analyses of next-generation sequencing (NGS) experiments, such as RNA-seq, often produce long lists of genes as output, such as those differentially expressed between two or more conditions. It is therefore a logical and critical next step to identify the biological relevance PLOS ONE | https://doi.org/10.1371/journal.pone.0272166 January 11, 2023 1 / 17 PLOS ONE and the NIH (R01 AI163742). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. MAGNET: Macrophage annotation of gene network enrichment tool associated with these genes. Typically, this is achieved through functional enrichment analyses that utilize standardized biological knowledge repositories, including Gene Ontology (GO) [1], Kyoto Encyclopedia of Genes and Genomes Pathways (KEGG) [2] and Molecular Signatures Database (MsigDB) [3]. These repositories annotate sets of genes with defined biological terms. These biological terms are then associated with input gene lists by calculating the overlap and performing statistical enrichment tests to assess significance. A large number of computational tools have been developed for performing this type of enrichment analysis by querying these repositories. Some of the most popular tools includes GOrilla [4], DAVID [5,6], IPA [7], and BiNGO [8]. Although these applications provide an effective way to characterize user-supplied gene lists and are now considered an essential step in typical bioinformatic workflows, they are often limited to generic results that do offer new insights to domain-specific questions. There are several reasons for this limitation: the attempt to provide terms that encompass all of biology; the static nature of the source repositories that do not account for the latest research; and the broad design of the annotation scheme. These issues are exemplified when endeavoring to perform gene set enrichment analysis on the results of genomic experiments using macrophages. Macrophages are highly plastic immune cells that are found in virtually every tissue in health as well as having an essential role in in innate immune response [9,10]. They exhibit highly specialized functions, depending on their tissue of residence and exhibit divergent responses to environmental and pathogenic stimuli [11,12]. They have been implicated in numerous pathological models and are under active investigation as potential therapeutic targets in various diseases [13]. For this reason, their genomic landscape has been the subject of many studies across multiple biomedical disciplines [14]. However, in our experience, typical gene set enrichment analysis tools using standardized repositories with the whole genome as background primarily return generic terms related to the role of macrophages in immune response and inflammation, regardless of the context of the original experiment. Alternatively, when a set of non-differentially expressed genes is included as background to account for the macrophage transcriptome, the tools may return no significant results at all. This is because the standardized repositories do not include annotation terms to describe the novel and specialized functions of macrophages. Instead, many canonical macrophage genes are associated with the role of macrophages in innate immunity despite their relevance to other biological processes and gene pathways. The plasticity of macrophages exacerbates this limitation, but the same issue arises across domains in studies that investigate the condition-sp (...truncated)