TSUNAMI: Translational Bioinformatics Tool Suite For Network Analysis And Mining

Zhi Huang, Zhi Han, Tongxin Wang, Wei Shao, Shunian Xiang, Paul Salama, Maher Rizkalla, Kun Huang, Jie ZhangbioRxiv 2019 · DOI

TL;DR

This paper introduces TSUNAMI, a comprehensive web-based tool suite for analyzing gene co-expression networks (GCNs). It addresses the need for an easy-to-use platform that integrates GCN mining with downstream analyses like Gene Ontology (GO) enrichment and survival analysis, making complex bioinformatics accessible to a wider range of researchers.

What Problem Does This Paper Address?

Understanding how genes work together is crucial for deciphering biological processes and identifying disease biomarkers. Gene co-expression network (GCN) analysis is a powerful approach to uncover these relationships by identifying groups of genes that are expressed similarly across different conditions or samples. However, existing tools often require significant bioinformatics expertise, lack integration of multiple analysis steps, or produce modules that are too large to be easily interpreted. Specifically, researchers need a user-friendly tool that can:

  • Mine GCN modules that are relatively small and tightly connected, facilitating downstream analyses.
  • Handle various types of transcriptomic data (microarray, RNA-seq, single-cell RNA-seq).
  • Integrate GCN mining with essential downstream analyses like Gene Ontology (GO) enrichment and survival analysis.
  • Provide visualization tools for better interpretation of results.

Methods

TSUNAMI offers a comprehensive pipeline for GCN analysis, incorporating two main algorithms:

  • lmQCM (imQCM): This is the core algorithm developed by the authors, designed to mine smaller, densely co-expressed GCN modules that can have overlapping gene memberships. This feature is highlighted as reflecting biological reality where genes participate in multiple pathways.
  • WGCNA (Weighted Gene Co-expression Network Analysis): A popular and established method, also integrated into TSUNAMI, allowing users to compare results from different approaches.

The tool provides a user-friendly web interface with flexible parameter selection for both algorithms. Key features include:

  • Data Input: Users can upload their own data (microarray, RNA-seq, etc.) or directly access data from public databases like NCBI GEO and TCGA.
  • Data Pre-processing: Includes steps for converting probe IDs to gene symbols, handling missing values, and filtering genes based on expression and variance.
  • GCN Mining: Implements both lmQCM and WGCNA for module identification.
  • Downstream Analysis:
    • GO Enrichment: Integrates with Enrichr to perform GO and pathway enrichment analysis.
  • Circos Plots: Visualizes gene locations on chromosomes and gene-gene relationships.
  • Survival Analysis: Correlates GCN module eigengenes with patient survival data (Overall Survival/Event-Free Survival).
  • Output: Results are downloadable in various formats (CSV, TXT), and visualizations are provided.

Datasets

The paper uses the GSE17537 microarray dataset from the Gene Expression Omnibus (GEO) as a primary example to demonstrate TSUNAMI's capabilities. This dataset comprises microarray data from 55 colorectal cancer patients. The authors also mention the integration of TCGA RNA-seq data and the ability to process data from NCBI GEO, which contains a large number of microarray datasets. The tool is designed to handle various numerical omics data types, including microarray, RNA-seq, scRNA-seq, and DNA methylation data.

Key Insights and Ideas

  • Bridging the Gap: TSUNAMI effectively bridges the gap between raw omics data and actionable biological insights by providing a unified, user-friendly platform for the entire analysis pipeline.
  • lmQCM's Advantage: The lmQCM algorithm's ability to identify smaller, potentially overlapping modules is highlighted as a key advantage for biological interpretability and downstream analysis, offering a complementary perspective to larger modules often generated by WGCNA.
  • Integrated Workflow: The seamless integration of GCN mining with GO enrichment and survival analysis is a significant contribution, reducing the need for users to manually combine results from disparate tools.
  • Accessibility: By offering a web server and clear parameter options, TSUNAMI democratizes GCN analysis, making it accessible to researchers without extensive programming or bioinformatics backgrounds.
  • Visualization Power: The inclusion of Circos plots for visualizing gene locations and relationships adds a powerful layer of interpretation for GCN modules.

Key Results

The paper demonstrates TSUNAMI's functionality using the GSE17537 dataset. For example, the 36th GCN module identified by lmQCM was found to be highly enriched with the GO Biological Process term "type I interferon signaling pathway." Furthermore, the survival analysis on this module showed a statistically significant difference in survival between high-risk and low-risk groups (p-value: 0.037613), suggesting its potential as a prognostic biomarker. The tool successfully processed this dataset, demonstrating its ability to handle real-world biological data.

Limitations and Future Work

  • Future Version 2.0: The authors plan to extend TSUNAMI to accept adjacency matrices with numerical values as edge weights, enabling the mining of a broader range of correlational network modules.
  • Python Upgrade: Future upgrades will incorporate Python to potentially improve the computing speed of the module mining step.
  • Species Support: While currently focused on human genomes for Circos plots, future work aims to incorporate genomes from other species, such as mice.
  • Customization: While the current version offers survival analysis as an example, future versions will allow for more customization of this feature to correlate module eigengenes with other clinical traits.

Summary updated

Abstract

Abstract Gene co-expression network (GCN) mining identifies gene modules with highly correlated expression profiles across samples/conditions. It helps to discover latent gene/molecular interactions, identify novel gene functions, and extract molecular features from certain disease/condition groups, thus help to identify disease biomarkers. However, there lacks an easy-to-use tool package for users to mine GCN modules that are relatively small in size with tightly connected genes that can be convenient for downstream Gene Ontology (GO) enrichment analysis, as well as modules that may share common members. To address this need, we develop a GCN mining tool package TSUNAMI (Tools SUite for Network Analysis and MIning) which incorporates our state-of-the-art lmQCM algorithm to mine GCN modules in public and user-input data (microarray, RNA-seq, or any other numerical omics data), then performs downstream GO and enrichment analysis based on the modules identified. It has several features and advantages: (i) user friendly interface and the real-time co-expression network mining through web server; (ii) direct access and search of GEO and TCGA databases as well as user-input expression matrix (microarray, RNA-seq, etc.) for GCN module mining; (iii) multiple co-expression analysis tools to choose with highly flexible of parameter selection options; (iv) identified GCN modules are summarized to eigengenes, which are convenient for user to check their correlation with other clinical traits; (v) integrated downstream Enrichr enrichment analysis and links to other GO tools; (vi) visualization of gene loci by Circos plot in any step. The web service is freely accessible through URL: http://spore.ph.iu.edu:3838/zhihuan/TSUNAMI/ . Source code is available at https://github.com/huangzhii/TSUNAMI/ .
BibTeX
@misc{huang2019tsunami,
  author = {Zhi Huang and Zhi Han and Tongxin Wang and Wei Shao and Shunian Xiang and Paul Salama and Maher Rizkalla and Kun Huang and Jie Zhang},
  title = {{TSUNAMI: Translational Bioinformatics Tool Suite For Network Analysis And Mining}},
  year = {2019},
  doi = {10.1101/787507},
  howpublished = {bioRxiv},
}
Figure