Hiplot: a comprehensive and easy-to-use web service boosting publication-ready biomedical data visualization

Jianfeng Li, Ben‐Ben Miao, Shixiang Wang, Wei Dong, Houshi Xu, Chenchen Si, Wei Wang, Songqi Duan, Jiacheng Lou, Zhiwei Bao, Hailuan Zeng, Zengzeng Yang, Wenyan Cheng, Fei Zhao, Jianming Zeng, Xuesong Liu, Renxie Wu, Yang Shen, Chen Zhu, Sai‐Juan Chen, Mingjie WangbioRxiv 2022 · DOI

TL;DR

Hiplot is a new, comprehensive, and user-friendly web service that simplifies the creation of publication-ready visualizations for complex biomedical data. It addresses the current gap in omics-focused platforms by offering over 230 interactive visualization tools, making advanced data analysis accessible to researchers without extensive programming skills.

What Problem Does This Paper Address?

Biomedical research generates vast amounts of complex data from clinical studies, omics experiments, and other sources. Visualizing this data is crucial for interpretation, reproducibility, and discovery. However, existing web-based bioinformatics services often fall short in providing robust, publication-quality visualization tools, especially for omics data. While some platforms offer basic visualization plugins, they lack the depth, interactivity, and ease of use required by biologists and clinicians who may not have strong bioinformatics or biostatistics backgrounds. This creates a bottleneck for researchers trying to generate compelling figures for their publications and conduct interactive data mining.

Methods

Hiplot tackles this challenge by offering a comprehensive suite of over 230 visualization plugins built on modern web technologies. Key methodological aspects include:

  • User-Friendly Interface: Developed using Vue.js and Vuetify.js, Hiplot provides a consistent and intuitive user experience with a two-column layout for easy navigation and interaction.
  • Interactive Spreadsheet Editor: A novel feature for large-scale biomedical visualization services, allowing users to preview, import, and edit tabular data directly within the web interface.
  • JSON-Based Plugin System: This system enables dynamic deployment of new plugins and supports both the web client and a cross-platform command-line interface (CLI) program called Hctl. Plugins are structured with Meta JSON (metadata), UI JSON (interface components), and Data JSON (default parameters and demo data).
  • Plumber-Based Task Response: For R-based web plugins, a plumber-based strategy is employed to significantly reduce the time for generating basic graphics by pre-loading dependencies and environments, allowing for near-instantaneous results for many tasks.
  • Cross-Platform CLI (Hctl): A Go-based command-line tool that allows users to perform visualizations in different environments, enhancing reproducibility and workflow integration.
  • Reproducibility Features: Users can conveniently reproduce past analyses by saving and loading task parameters as JSON files, and the system provides logging for debugging.
  • Scalable Infrastructure: Built on high-performance computing nodes with substantial storage and memory, Hiplot can handle large datasets and numerous concurrent users.

Datasets

The paper demonstrates Hiplot's functionality using both demo datasets and real-world biomedical data. Specific examples mentioned include:

  • Iris dataset: Used to showcase dimensionality reduction visualization.
  • Lung dataset: Utilized for demonstrating predictive risk scores in survival analysis.
  • TCGA (The Cancer Genome Atlas) database: Used for exploring gene expression and mutation data in various cancer types.
  • UCSC Xena data hubs: Data from these hubs are integrated and visualized through applications like UCSCXenaShiny.
  • BCP-ALL (B-cell precursor acute lymphoblastic leukemia) variants data: Used to validate fusion gene and chromosomal abnormality analyses.

The paper does not introduce new benchmarks but rather leverages existing public datasets to showcase the capabilities of Hiplot.

Key Insights and Ideas

  • Democratizing Advanced Visualization: Hiplot significantly lowers the barrier to entry for complex data visualization, empowering researchers who lack specialized programming skills to generate publication-quality figures.
  • Comprehensive Functionality: The sheer number and diversity of plugins (over 230) covering basic graphics, omics, clinical data, and machine learning approaches make Hiplot a one-stop shop for many visualization needs.
  • Integrated Workflow: The combination of a user-friendly web interface, an interactive spreadsheet editor, and a powerful CLI (Hctl) provides flexibility for different user preferences and integration into existing pipelines.
  • Speed and Efficiency: The plumber-based response strategy for R plugins dramatically speeds up the generation of basic graphics, a critical factor for interactive data exploration.
  • Community-Driven Development: The Hiplot Consortium's collaborative approach fosters continuous development and expansion of the platform's capabilities.

Key Results

  • Extensive Plugin Library: Hiplot offers over 230 interactive visualization plugins, significantly more than comparable services like ImageGP (16) and Galaxy (60+).
  • High User Engagement: Since its launch, Hiplot has attracted over 2.5 million visits from 100 countries, with an average of 5,000 visits and 3,000 task submissions per day. Over 22,000 user accounts have been registered.
  • Popularity of Key Plugins: Plugins like "heatmap" have been visited over 70,000 times, indicating strong demand for core visualization functionalities.
  • Comparable to Commercial Software: The quality of visualizations is noted to be comparable to commercial desktop software like GraphPad.
  • Reduced Task Time: The plumber-based strategy allows basic graphics to be completed in seconds, a significant improvement over traditional methods.

Limitations and Future Work

  • Workflow Editor Support: The paper notes that the setup of tasks using a workflow editor is not yet supported, though it is considered feasible for future integration.
  • Shiny/Streamlit Plugin Error Reproduction: The authors acknowledge that reproducing errors in Shiny- and Streamlite-based plugins remains challenging.
  • Future Integration: The authors express interest in building a Galaxy-like interface for better workflow-based integration if there is strong user demand.
  • Continuous Development: The website is under daily updates, indicating an ongoing development cycle with potential for further enhancements and new plugins.

Summary updated

Abstract

Abstract Modern web techniques provide an unprecedented opportunity for leveraging complex biomedical data generating in clinical, omics, and mechanism experiments. Currently, the functions for carrying out publication-ready biomedical data visualization represent primary technical hurdles in the state-of-art omics-based web services, whereas the demand for visualization-based interactive data mining is ever-growing. Here, we propose an easy-to-use web service, Hiplot ( https://hiplot.com.cn ), equipping with comprehensive and interactive biomedical data visualization functions (230+) including basic statistics, multi-omics, regression, clustering, dimensional reduction, meta-analysis, survival analysis, risk modeling, etc. We used the demo and real datasets to demonstrate the usage workflow and the core functions of Hiplot. It permits users to conveniently and interactively complete a few specialized visualization tasks that previously could only be done by senior bioinformatics or biostatistics researchers. A modern web client with efficient user interfaces and interaction methods has been implemented based on the custom components library and the extensible plugin system. The versatile output can also be produced in different environments via using the cross-platform portable command-line interface (CLI) program, Hctl. A switchable view between the editable data table and the file uploader/path selection could facilitate data importing, previewing, and exporting, while the plumber-based response strategy significantly reduced the time costs for generating basic scientific graphics. Diversified layouts, themes/styles, and color palettes in this website allow users to create high-quality and publication-ready graphics. Researchers devoted to both life and data science may benefit from the emerging web service.
BibTeX
@misc{li2022hiplot,
  author = {Jianfeng Li and Ben‐Ben Miao and Shixiang Wang and Wei Dong and Houshi Xu and Chenchen Si and Wei Wang and Songqi Duan and Jiacheng Lou and Zhiwei Bao and Hailuan Zeng and Zengzeng Yang and Wenyan Cheng and Fei Zhao and Jianming Zeng and Xuesong Liu and Renxie Wu and Yang Shen and Chen Zhu and Sai‐Juan Chen and Mingjie Wang},
  title = {{Hiplot: a comprehensive and easy-to-use web service boosting publication-ready biomedical data visualization}},
  year = {2022},
  doi = {10.1101/2022.03.16.484681},
  howpublished = {bioRxiv},
}
Figure