mCNA

mCNA analyzes copy number variations (CNVs) from next-generation sequencing (NGS) data using unique molecular identifiers (UMIs) to improve the accuracy and resolution of CNV detection in tumor genomes.

Key Features:

• UMI-based labeling: Employs unique molecular identifiers (UMIs), short random nucleotide sequences appended to sequencing primers, to uniquely label DNA molecules and reduce technical artifacts.
• UMI count matrices: Constructs count matrices that catalog the frequency of each UMI across sequencing reads.
• Pseudo-reference construction: Builds a pseudo-reference genome from control samples for baseline comparison and normalization.
• Log-ratio computation: Calculates log-ratios relative to the pseudo-reference to quantify copy number deviations.
• Segmentation and statistical inference: Segments the genome by consistent log-ratio values and applies statistical inference to identify CNV breakpoints.
• NGS compatibility: Operates on next-generation sequencing (NGS) data.
• Validation: Validated on a Diffuse Large B-cell Lymphoma patient dataset with strong correlation to comparative genomic hybridization results.

Scientific Applications:

• High-resolution CNV detection in cancer: Detects copy number gains, amplifications, and deletions in tumor genomes at high resolution.
• Artifact discrimination: Distinguishes true somatic CNVs from technical artifacts using UMI-based molecule identification.
• Mechanistic studies: Enables investigation of oncogene activation and tumor suppressor gene inactivation driven by CNVs.
• Cross-platform comparison: Supports comparative analysis of NGS-derived CNV profiles against comparative genomic hybridization.
• Disease-specific analysis: Applied to Diffuse Large B-cell Lymphoma datasets for CNV discovery and validation.

Methodology:

The computational workflow comprises construction of UMI count matrices, pseudo-reference construction using control samples, computation of log-ratios versus the pseudo-reference, and genomic segmentation followed by statistical inference to call CNVs.