DECOMICS, a shiny application for unsupervised cell type deconvolution and biological interpretation of bulk omic...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Estimating cell composition from bulk tissue samples using unsupervised deconvolution algorithms is challenging, particularly for researchers without bioinformatics expertise.
• Interpreting the results of cell-type deconvolution requires specialized knowledge and tools.

Purpose of the Study:

• To develop and present DECOMICS, a user-friendly shiny-web application for unsupervised deconvolution of bulk omic data.
• To provide tools for estimating and identifying cell type composition from transcriptomic and methylomic data.

Main Methods:

• DECOMICS is an R-shiny web application implementing various unsupervised deconvolution algorithms for gene expression or methylation data.
• It includes an enrichment analysis module for biological interpretation of deconvolved components and visualization tools.
• The application supports data preprocessing, including normalization, transformation, and feature selection.

Main Results:

• DECOMICS offers a comprehensive platform for unsupervised deconvolution, integrating multiple algorithms and analytical modules.
• The tool facilitates the estimation and identification of cell type composition from bulk omic datasets.
• Results from deconvolution, enrichment analysis, and visualization are downloadable for further use.

Conclusions:

• DECOMICS lowers the barrier to entry for cell-type deconvolution analysis, empowering researchers to interpret bulk omic data.
• The application enhances the biological interpretation of deconvolved cell components through integrated enrichment analysis.
• DECOMICS is available as an R package and on the Biosphere-IFB Clouds Federation for scalable high-performance computing.