Characterizing allele- and haplotype-specific copy numbers in single cells with CHISEL

Area of Science:

• Genomics
• Cancer Biology
• Computational Biology

Background:

• Single-cell sequencing offers parallel genome analysis but suffers from very low coverage (<0.05×).
• Existing methods can determine large segment copy numbers but miss critical allele-specific mutations, such as copy-neutral loss of heterozygosity (LOH).

Purpose of the Study:

• To introduce CHISEL (copy-number haplotype inference in single cells using evolutionary links).
• To infer allele- and haplotype-specific copy numbers from sparse single-cell sequencing data.
• To analyze allele-specific copy-number aberrations (CNAs) in breast cancer.

Main Methods:

• Developed CHISEL, a computational method aggregating sparse signals across thousands of single cells.
• Applied CHISEL to ten single-cell sequencing datasets from two breast cancer patients (~2,000 cells total).

Main Results:

• Identified extensive allele-specific CNAs, including copy-neutral LOHs, whole-genome duplications (WGDs), and mirrored-subclonal CNAs.
• Observed that these CNAs impact genomic regions with known breast cancer genes.
• Refined tumor evolution reconstruction, timing CNAs relative to WGDs and identifying low-frequency subpopulations.

Conclusions:

• CHISEL effectively infers allele-specific copy numbers from low-coverage single-cell data.
• The method uncovers novel insights into breast cancer heterogeneity and evolution, including convergent evolution.
• CHISEL enhances the resolution of genomic analysis in single-cell studies.