MonaGO: a novel Gene Ontology enrichment analysis visualisation system

Ziyin Xin, Yujun Cai, Louis T. Dang, Hannah Burke, Jerico Revote, Hieu T. Nim, Yuan-Fang Li, Mirana RamialisonbioRxiv 2020 · DOI

TL;DR

This paper introduces MonaGO, a novel web-based system for visualizing Gene Ontology (GO) enrichment analysis results. MonaGO offers an intuitive, interactive, and customizable interface that allows biologists to explore and interpret complex GO data more effectively, addressing limitations in existing tools by providing a unified, dynamic visualization experience.

What Problem Does This Paper Address?

Gene Ontology (GO) enrichment analysis is a crucial step in post-omics research, helping scientists understand the biological functions of large gene sets. However, interpreting the results can be challenging due to the sheer volume of data and the hierarchical nature of GO terms, which can lead to redundancy and difficulty in identifying truly representative terms. Existing tools often present results as static tables or simple graphs, or require users to navigate multiple pages and tools, hindering efficient exploration and interpretation. This can lead to biased selection of terms or missed biological insights. MonaGO aims to solve this by providing a single, interactive platform that integrates analysis, visualization, and exploration.

Methods

MonaGO employs a client-server architecture. The backend, built with Python and Flask, handles data processing, including calculating similarities between GO terms using various metrics (Jaccard similarity based on overlapping genes, Resnik semantic similarity, and SimRel semantic similarity). It also implements hierarchical clustering algorithms to group similar GO terms. The frontend, developed with JavaScript and D3.js, provides a dynamic and interactive chord diagram visualization. This visualization allows users to explore clusters, view gene overlap, and access detailed information about GO terms. Key features include:

  • Interactive Chord Diagram: Visualizes GO term clusters and their relationships.
  • Dynamic Clustering: Users can systematically or manually cluster GO terms.
  • Multiple Similarity Metrics: Supports gene overlap and semantic similarity (Resnik, SimRel) with options for average, minimum, or maximum cluster comparison.
  • Data Entry Flexibility: Accepts gene lists (integrating with DAVID for enrichment) or pre-selected GO terms.
  • Reproducibility: Allows saving and restoring visualization sessions.
  • Export Options: High-resolution images (PDF, PNG, SVG) and session data (JSON).

Datasets

The paper describes using a curated list of zebrafish embryonic cardiac genes for evaluating MonaGO's performance. This dataset was used to assess the biological outcomes of clustering using different similarity measures.

Key Insights and Ideas

  • Unified Visualization: MonaGO's core innovation is presenting interactive GO enrichment results and their hierarchical clustering within a single output page, a feature not found in other compared tools.
  • Bridging Gene Overlap and Semantic Similarity: The system effectively combines gene overlap and semantic similarity measures for clustering, allowing for the discovery of functionally related terms even when direct gene overlap is absent. This is crucial for uncovering novel biological hypotheses.
  • Reducing Information Overload: Hierarchical clustering and the chord diagram visualization help reduce redundancy and present complex GO data in a more digestible and interpretable format for biologists.
  • Facilitating Hypothesis Generation: By enabling interactive exploration and revealing functional links between terms, MonaGO empowers biologists to explore novel hypotheses that might be overlooked with traditional methods.

Key Results

A comparison with 11 other GO enrichment visualization tools revealed that MonaGO is unique in its ability to offer simultaneous interactive visualization within a single output page. The system successfully demonstrated its utility in analyzing zebrafish embryonic cardiac genes, identifying functional links between liver, thyroid, and heart development, and suggesting novel hypotheses about gene involvement in heart left-right asymmetry. The interactive features, such as manual clustering and detailed gene information display, were highlighted as significant advantages.

Limitations and Future Work

The paper does not explicitly detail limitations or future work beyond the general statement that MonaGO aims to provide a unique tool for biologists interested in hands-on interaction with gene lists and their semantic relationships. However, the reliance on DAVID for initial enrichment analysis for gene lists could be seen as a dependency. Future work might involve expanding the integrated enrichment analysis capabilities or further refining the clustering algorithms and visualization options.

Summary updated

Abstract

Abstract MonaGO is a novel web-based visualisation system that provides an intuitive, interactive and responsive interface for performing gene ontology (GO) enrichment analysis and visualising the results. MonaGO combines dynamic clustering and interactive visualisation as well as customisation options to assist biologists in obtaining meaningful representation of overrepresented GO terms, producing simplified outputs in an unbiased manner. MonaGO supports gene lists as well as GO terms as inputs. Visualisation results can be exported as high-resolution images or restored in new sessions, allowing reproducibility of the analysis. An extensive comparison between MonaGO and 11 state-of-the-art GO enrichment visualisation tools based on 9 features revealed that MonaGO is the only platform that simultaneously allows interactive visualisation within one single output page, directly accessible through a web browser with customisable display options. In summary, MonaGO will facilitate the interpretation of GO analysis and will assist the biologists into the representation of the results.
BibTeX
@misc{xin2020monago,
  author = {Ziyin Xin and Yujun Cai and Louis T. Dang and Hannah Burke and Jerico Revote and Hieu T. Nim and Yuan-Fang Li and Mirana Ramialison},
  title = {{MonaGO: a novel Gene Ontology enrichment analysis visualisation system}},
  year = {2020},
  doi = {10.1101/2020.09.27.316067},
  howpublished = {bioRxiv},
}
Figure