ClusterApp to visualize, organize, and navigate metabolomics data

Area of Science:

• Metabolomics
• Bioinformatics
• Data Visualization

Background:

• Clustering analysis is crucial for exploratory data analysis, but traditional methods like Principal Component Analysis (PCA) are often suboptimal for metabolomics data visualization.
• Metabolomics datasets require specialized preprocessing, and a lack of user-friendly tools hinders researchers without computational expertise.
• ClusterApp provides an accessible web application for Principal Coordinate Analysis (PCoA), expanding clustering options in metabolomics.

Purpose of the Study:

• To introduce ClusterApp, a novel web application designed to simplify clustering analysis and data visualization in metabolomics.
• To provide metabolomics researchers with an accessible tool for exploratory data analysis, reducing the need for advanced bioinformatics skills.
• To offer flexible data input and analysis options, including integration with existing metabolomics data formats and platforms.

Main Methods:

• ClusterApp utilizes Principal Coordinate Analysis (PCoA) for clustering and visualization, built upon a QIIME 2 Docker image.
• The application supports data input from GNPS, GNPS2, or user-provided spreadsheets, offering flexibility in data integration.
• It incorporates data preprocessing techniques such as blank removal, Total Ion Current (TIC) normalization, auto-scaling, and targeted filtering.

Main Results:

• Analysis of LC-MS/MS metabolomics datasets (mouse tissue and coral life stages) demonstrated ClusterApp's effectiveness in revealing biological variations.
• The Bray-Curtis dissimilarity measure, combined with targeted filtering and TIC normalization with auto-scaling, significantly enhanced data reliability and clustering resolution.
• ClusterApp's features allowed for tailored analyses, improving the visualization and interpretation of metabolomic profiles, highlighting distinct separations in sample groups.

Conclusions:

• ClusterApp provides an accessible and dynamic solution for exploratory data analysis in metabolomics, addressing the need for user-friendly visualization tools.
• By integrating data transformation capabilities with PCoA, ClusterApp offers a versatile platform for clustering analysis, suitable for various research needs.
• The application's web interface and Docker-based deployment empower researchers to uncover patterns in metabolomics data without requiring extensive computational expertise or complex data manipulation.