Crambled: A Shiny application to enable intuitive resolution of conflicting cellularity estimates

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Whole-genome sequencing is standard for tumor sample analysis.
• Common tasks include estimating tumor cellularity, profiling copy numbers, and assessing sub-clonal evolution.
• Existing analysis tools often yield conflicting results due to inherent model uncertainties.

Purpose of the Study:

• To introduce "Crambled", a novel R Shiny application.
• To enable intuitive visual comparison of conflicting solutions from genomic analysis tools.
• To aid researchers in understanding and selecting appropriate results for tumor genomic profiling.

Main Methods:

• Development of "Crambled" as an R Shiny application.
• Implementation of functions to generate visualizations from sequencing data.
• Inclusion of example use cases with tumor and cell line data.

Main Results:

• "Crambled" facilitates visual comparison of divergent analytical outcomes.
• The tool helps elucidate the reasons behind differing tool recommendations.
• Users can make more informed decisions regarding competing genomic profiles.

Conclusions:

• "Crambled" addresses the challenge of conflicting results in cancer genomics.
• Visual comparison aids in understanding and resolving analytical discrepancies.
• The application supports informed interpretation of tumor purity and copy number data.