QDNAseq

QDNAseq detects DNA copy number aberrations from shallow whole-genome sequencing (WGS) data and corrects technical and reference biases to enable copy-number profiling of low-quality samples such as formalin-fixed paraffin-embedded (FFPE) tissue.

Key Features:

• Low-Coverage Efficiency: Performs copy-number profiling from shallow WGS at approximately 0.1× genome coverage.
• Correction for Mappability and GC Content: Implements a combined correction for sequence mappability and GC content using a two-dimensional loess approach.
• Blacklist Development: Derives a genome blacklist of problematic regions by leveraging data from the 1000 Genomes Project to filter spurious signals, with some overlap to ENCODE annotations.
• Bias Correction: Accounts for errors in the human reference genome, repetitive sequences, polymorphisms, variable sample quality, and sequencing-introduced biases.
• Robustness for Degraded DNA: Optimized to handle degraded DNA from FFPE samples and reduce FFPE-specific artifacts.
• Integration with Established Packages: Uses DNAcopy for segmentation and CGHcall for copy-number calling as downstream analysis steps.

Scientific Applications:

• Cancer research: Facilitates detection and study of somatic copy-number alterations in archival tumor tissue, including FFPE samples.
• Large-scale profiling: Enables generation of high-quality genome profiles with low noise near the read-counting statistical limit and has been applied to over 1,000 samples from more than 25 institutions.

Methodology:

Processes shallow WGS data (~0.1× coverage), applies two-dimensional loess correction for mappability and GC content, filters regions using a 1000 Genomes Project-derived blacklist, segments with DNAcopy, and calls CNAs with CGHcall.