Addressing multiple facets of ligand-receptor network inference including single-cell proteomics

Area of Science:

• Computational Biology
• Bioinformatics
• Systems Biology

Background:

• Ligand-receptor interaction (LRI) inference tools yield inconsistent results due to dataset variations.
• Existing tools struggle with diverse experimental designs and data features, hindering universal applicability.

Purpose of the Study:

• To develop an integrated computational framework for robust LRI inference.
• To enhance the SingleCellSignalR package with flexible scoring and analytical depth options.
• To enable cross-modal LRI comparisons using single-cell transcriptomics and proteomics.

Main Methods:

• Expanded the SingleCellSignalR Bioconductor package to include adjustable analytical depth and alternative scoring strategies.
• Analyzed two single-cell transcriptomics datasets with contrasting experimental designs.
• Performed a direct comparison of LRI inference between single-cell proteomics and transcriptomics data.
• Integrated additional data types, including patient-derived mouse xenografts and bulk RNA sequencing.

Main Results:

• The enhanced framework demonstrates flexibility in handling diverse single-cell datasets.
• A novel comparison of LRI inference across single-cell proteomics and transcriptomics modalities was achieved.
• The framework successfully accommodates various data types, expanding LRI analysis capabilities.

Conclusions:

• The integrated SingleCellSignalR framework offers a versatile solution for LRI inference across multiple single-cell data types.
• This approach facilitates robust and comparable LRI analysis, addressing limitations of existing tools.
• The framework supports novel cross-modal analyses and integrates emerging data types for broader biological insights.