RadTrixVis: A Dashboard for Visualizing An Unbalanced Bipartite Graph

If you use this code in your work, please cite it as follows:

Reddy Rani Vangimalla, and J. Sreevalsan-Nair, "RadTrix: A Composite Hybrid Visualization for Unbalanced Bipartite Graphs in Biological Datasets," 9th Eurographics Workshop on Visual Computing for Biology and Medicine, September 2019 (VCBM 2019). Conference proceedings.

Set Up and Running the Project

1. Clone the Repository

   • If you haven't cloned it yet:

     git clone <repository_url>

   • If already cloned:

     cd <project_folder>
     git pull origin main  # Replace 'main' with the correct branch if needed

2. Install Dependencies

   npm install  # Use 'yarn' if preferred

3. Start the Development Server

   npm start  # Use 'yarn start' if using Yarn

   This launches the app at http://localhost:3000/.

4. Troubleshooting (if needed)
   If you encounter issues after npm install, try:

   rm -rf node_modules
   npm cache clean --force
   npm install

Your app should now be running locally! 🚀

To launch the dashboard

https://gvcl.github.io/RadTrixVis/

RadTrixVis GUI Tool Help Manual

About

RadTrixVis is a powerful GUI tool designed for analyzing gene-phenotype relationships through phenotype-specific filtering. It enables users to identify significant correlations and network interactions within biological data efficiently.

---

Key Features

• Phenotype-Specific Filtering: Three filtering modes—Correlation-Based, Network-Based, and New Data Input—ensure up-to-date visualizations.
• Disease Phenotype Section: Provides phenotype selection, search functionality, and color coding for enhanced clarity.
• Gene Section Management: Features hidden and shown gene lists, search capabilities, and bulk selection actions.
• Top K Genes & Customized Gene Selection: Allows prioritization and filtering of genes based on ranking and selection criteria.
• Download Data Panel: Exports data and visuals in JSON, CSV, and image formats.
• Aesthetics Panel: Customizes color schemes and node ordering to improve visualization clarity.

---

Phenotype-Specific Filtering

RadTrixVis provides three different modes for filtering genes based on phenotype relationships:

1. Correlation-Based Filtering

   • Identifies and filters genes statistically correlated with a given phenotype.
   • Uses correlation metrics to determine gene relevance.

2. Network-Based Filtering

   • Analyzes genes based on their roles within biological networks.
   • Uses connectivity and network-based algorithms for filtering.

3. New Input Data Filtering

   • Allows users to upload and analyze custom datasets for tailored phenotype investigations.
   • Supports various data formats for input.

---

Using the Update Data Button

1. Select "New Input Data" mode.
2. Upload a compatible dataset file.
3. Click "Update Data" to refresh the analysis visuals with the new dataset.

---

Disease Phenotype Section

• Dropdown Selection: Users can include/exclude specific phenotypes.
• Color-Coded Tags: Helps visually distinguish different phenotypes for easier analysis.

---

Managing Gene Visibility

• RadTrixVis provides two lists for genes:
  • Shown Genes: Genes currently included in the analysis.
  • Hidden Genes: Genes excluded but still available for selection.
• Search & Bulk Selection:
  • Use the search function to find specific genes quickly.
  • Checkboxes allow bulk actions to show or hide multiple genes at once.

---

Top K Genes Inclusion

• Users can specify the number of top-ranked genes to focus on.
• Selection is based on ranking metrics relevant to the phenotype analysis.

---

Customized Gene Selection

• Allows rank-based filtering from a range of specified ranks.
• Users can adjust step sizes to fine-tune gene selection granularity.

---

Data and Image Export Options

RadTrixVis supports multiple export formats to suit different analysis needs:

• JSON: Structured data export for further processing.
• CSV: Tabular format for spreadsheet and statistical tool compatibility.
• Image Formats: Export visual representations for reports and presentations.

---

Aesthetics Customization

Users can adjust visualization settings to enhance interpretability:

1. Phenotype & Gene Colors

   • Assign specific colors to different phenotypes and gene nodes.
   • Helps create distinct visual categories in network graphs.

2. Node Ordering

   • Order nodes based on:
     • Degree (importance in the network)
     • Random (for unbiased visualization)
     • Lexicographical (alphabetical sorting for structured layouts)

---

Summary of Steps

1. Choose Filtering Mode: Correlation-Based, Network-Based, or New Input Data.
2. Upload Data (if required) and click Update Data to refresh visuals.
3. Manage Phenotypes: Use dropdown selection and color-coded tags.
4. Control Gene Visibility: Use checkboxes, search, and bulk selection.
5. Select Top K Genes or Customized Gene Range for focused analysis.
6. Customize Aesthetics: Adjust colors and node ordering for better clarity.
7. Export Data & Visuals in JSON, CSV, or image formats.

---