CNAViz: An interactive webtool for user-guided segmentation of tumor DNA sequencing data

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Copy-number aberrations (CNAs) are crucial in cancer, but their identification from DNA sequencing data is challenging due to complex genomic segmentation.
• Current segmentation algorithms have limitations, either causing overclustering (local) or omitting focal CNAs (global), necessitating manual correction in pan-cancer studies.

Purpose of the Study:

• To introduce CNAViz, a web-based tool designed to enhance copy-number segmentation accuracy by integrating both local and global approaches.
• To overcome the limitations of existing segmentation methods and improve the reliability of CNA identification in cancer research.

Main Methods:

• CNAViz enables simultaneous local and global segmentation, addressing the shortcomings of individual approaches.
• The tool's performance was evaluated using simulated data and six bulk DNA sequencing samples from breast cancer patients.

Main Results:

• CNAViz demonstrated superior segmentation accuracy compared to existing local and global methods on simulated data.
• Validation with single-cell DNA sequencing data confirmed that CNAViz improves segmentation and downstream copy-number calling accuracy in breast cancer samples.

Conclusions:

• CNAViz offers a significant advancement in copy-number segmentation, providing a more accurate and efficient tool for cancer genomics research.
• The integrated local and global segmentation approach in CNAViz reduces the need for manual correction, streamlining analysis for researchers.