Vardetector: a pure Python package to detect DNA called mutations in aligned RNA reads

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Accurate detection of DNA mutations from RNA sequencing (RNA-seq) is crucial for cancer research.
• Formalin-fixed paraffin-embedded (FFPE) tumor samples present unique challenges for RNA-seq library preparation.
• Existing variant calling tools may have limitations in detecting mutations from RNA data.

Purpose of the Study:

• To introduce Vardetector, a freely available Python package for detecting DNA mutations in aligned RNA reads.
• To benchmark Vardetector against the industry-standard GATK HaplotypeCaller.
• To compare the efficacy of exome-capture versus ribosome-depletion RNA preparation protocols for FFPE tumor samples in mutation detection.

Main Methods:

• Development of the Vardetector Python package.
• Benchmarking Vardetector against GATK HaplotypeCaller using RNA-seq data.
• Calling somatic mutations from DNA of non-small cell lung cancer (NSCLC) tumor/normal samples.
• Comparison of exome-capture and ribodepletion RNA library preparation protocols by quantifying RNA reads over mutation sites.

Main Results:

• Vardetector demonstrates robust performance in detecting DNA mutations from RNA reads.
• The ribodepletion protocol significantly increases the number of all RNA reads and mutation-supporting reads compared to exome-capture (p < 0.001).
• Ribodepletion also provides significantly wider coverage breadth across somatic mutation positions (p < 0.001).

Conclusions:

• Vardetector is a valuable tool for DNA variant detection in RNA sequencing data.
• Ribosome-depletion RNA preparation is superior for FFPE tumor samples, yielding more comprehensive mutation data.
• These findings have significant implications for improving neoantigen prioritization pipelines in cancer research.