Next-generation deconvolution of the tumor microenvironment with omnideconv

Area of Science:

• Computational Biology
• Bioinformatics
• Cancer Research

Background:

• The tumor microenvironment, including immune cells, significantly impacts cancer progression and therapeutic outcomes.
• Cell-type deconvolution from bulk RNA sequencing (RNA-seq) estimates cellular composition using specific expression signatures.
• Emerging algorithms learn cell-type signatures from single-cell RNA sequencing (scRNA-seq) data, offering broader applicability but facing integration challenges due to diverse workflows.

Purpose of the Study:

• To address the complexity and heterogeneity of next-generation deconvolution algorithms.
• To introduce omnideconv, an R package designed to unify and simplify the usage of various deconvolution methods.
• To demonstrate the application of omnideconv for quantifying cellular composition in breast cancer bulk RNA-seq data using scRNA-seq references.

Main Methods:

• Development of the omnideconv R package, integrating multiple deconvolution algorithms.
• Unification of semantic input/output formats for diverse deconvolution tools.
• Integration with annotated scRNA-seq datasets, including malignant and normal breast cancer cells.

Main Results:

• Omnideconv successfully integrates several deconvolution methods within a single R package.
• The package streamlines the process of learning cell-type-specific signatures from scRNA-seq data.
• Demonstrated quantification of cellular composition in breast cancer patient cohorts using bulk RNA-seq data.

Conclusions:

• Omnideconv provides a unified and user-friendly platform for advanced cell-type deconvolution.
• Facilitates the analysis of complex cellular compositions within the tumor microenvironment.
• Enhances the utility of scRNA-seq data for interpreting bulk RNA-seq profiles in cancer research.