SinCHet

SinCHet quantifies single-cell heterogeneity in cancer by computing Shannon Profiles (SP) across clonal resolutions and comparing cell populations via the Profile of Shannon Difference (PSD) and its D statistic.

Key Features:

• Implementation: Implemented in MATLAB.
• Data types: Handles continuous data such as mRNA expression levels and binary omics data including discretized methylation profiles.
• Shannon Profile (SP): Quantifies cellular heterogeneity using the Shannon Profile evaluated across multiple clonal resolutions.
• Profile of Shannon Difference (PSD) and D statistic: Detects heterogeneity differences between two cell populations by computing the PSD and summarizing it via the D statistic as the area under the PSD.
• Clonal resolution detection: Determines a default clonal resolution by detecting change points in the PSD using a multivariate adaptive regression splines model (MARS).
• Custom clonal resolutions: Allows specification of custom clonal resolutions for targeted analyses.
• Biomarker prioritization: Supports prioritization of biomarkers based on heterogeneity or marker differences within and between cell populations.

Scientific Applications:

• Single-cell heterogeneity analysis: Analysis of cellular heterogeneity in cancer using mRNA expression and discretized methylation data.
• Population comparison and clonal dynamics: Identification of emerging or disappearing clones and comparison of heterogeneity between two cell populations.
• Biomarker selection and experimental planning: Prioritization of biomarkers and informing follow-up experiments based on heterogeneity and marker differences.

Methodology:

Computes Shannon Profile (SP) across clonal resolutions; computes Profile of Shannon Difference (PSD) between two populations; calculates the D statistic as the area under the PSD; detects change points in the PSD using a multivariate adaptive regression splines (MARS) model to set a default clonal resolution; accepts custom clonal resolutions; processes continuous (mRNA expression) and binary (discretized methylation) inputs; implemented in MATLAB.