katdetectr

katdetectr detects and characterizes kataegis loci and visualizes regional genomic hypermutation in whole-genome sequencing and other variant datasets.

Key Features:

• Detection and characterization: Identifies kataegis loci by locating regions with significantly elevated mutation rates relative to the mutational background.
• Data import flexibility: Accepts MAF (Mutation Annotation Format), VCF (Variant Call Format), and VRanges inputs for compatibility with common genomic variant data.
• Intermutation distance analysis: Calculates intermutation distances of genomic variants to identify clusters of closely spaced somatic mutations.
• Unsupervised changepoint analysis: Uses the Pruned Exact Linear Time (PELT) algorithm to detect shifts in mutation rates for kataegis calling with user-defined parameters.
• Visualization: Produces visual representations of identified kataegis loci and their mutational patterns.
• Standardized output: Exports results in a standardized format suitable for downstream analyses and integration into research workflows.

Scientific Applications:

• Cancer genomics: Detection and characterization of kataegis to study regional hypermutation across malignancies using whole-genome sequencing data.
• Tumor evolution and biomarker research: Analysis of kataegis patterns to inform studies of tumor evolution and potential therapeutic targets.

Methodology:

Imports MAF/VCF/VRanges files, computes intermutation distances, applies unsupervised changepoint analysis using the PELT algorithm for kataegis calling with user-defined parameters, and was evaluated on synthetic and a priori labeled pan-cancer whole-genome sequenced datasets against five other kataegis detection packages with reported accuracy and normalized Matthews Correlation Coefficient metrics.