CAISC

CAISC integrates copy number variation (CNV) and single nucleotide variation (SNV) genotype matrices from single-cell RNA sequencing (scRNA-seq) to cluster cells into subclones and characterize intratumor heterogeneity.

Key Features:

• CNV and SNV integration: Integrates copy number variation (CNV) and single nucleotide variation (SNV) genotype matrices derived from scRNA-seq.
• Entropy-weighted approach: Uses an entropy-weighted approach to combine CNV and SNV information.
• Subclone clustering: Clusters individual cells into distinct subclones based on the integrated genotype matrices.
• Improved assignment accuracy: Assigns subclones with increased accuracy compared to analyses using only one data type.
• Expression comparison: Enables analysis of expression changes between identified subclones.
• Validation: Evaluated on simulated data and four real-world datasets.
• MDS trisomy 8 example: Successfully delineated trisomy 8 clones in a myelodysplastic syndromes (MDS) dataset.
• Implementation: Implemented as an R package.

Scientific Applications:

• Intratumor heterogeneity analysis: Characterizes genetic heterogeneity within tumor samples by combining CNV and SNV data.
• Clonal group detection: Identifies and assigns cells to clonal groups (subclones) within tumors.
• Tumor evolution and clonal architecture: Supports reconstruction and interpretation of clonal architecture and evolutionary relationships.
• Differential expression between subclones: Detects expression changes associated with distinct subclones.
• Chromosomal alteration analysis: Facilitates detection and analysis of chromosomal events such as trisomy 8 in MDS.

Methodology:

Integrates CNV and SNV genotype matrices using an entropy-weighted approach to cluster cells into subclones; evaluated on simulated data and four real-world datasets; implemented in R.