scFeatures

scFeatures generates multi-view sample-level feature representations from single-cell and spatial data by constructing 17 distinct feature types grouped into six categories to enable interpretation of cellular heterogeneity and disease mechanisms.

Key Features:

• Multi-View Representations: Constructs diverse sample-level features from single-cell and spatial data to capture cellular and molecular characteristics.
• Seventeen Feature Types: Builds 17 distinct feature types categorized into six groups.
• Interpretable Summaries: Compresses complex cell-level data into interpretable summary statistics for each sample.
• Sample-Level Characterization: Aggregates cell-level measurements into summary features representing individual samples.
• Disease Mechanism Insights: Summarizes a broad collection of features to assist interpretation of underlying disease mechanisms across studies.
• Disease Classification: Produces comprehensive feature sets that support accurate classification of disease status at the individual sample level.

Scientific Applications:

• Oncology: Supports identification of cancer-related biomarkers and pathways using sample-level feature summaries.
• Immunology: Enables analysis of immune cell heterogeneity and immune-related molecular signatures.
• Developmental Biology: Facilitates study of cellular diversity and developmental processes through aggregated features.
• Biomarker Discovery: Aids identification of sample-level biomarkers by providing interpretable multiview features.
• Disease Pathway Elucidation: Assists elucidation of disease pathways via summarized feature collections.
• Therapeutic Strategy Design: Supports tailoring therapeutic strategies through molecular characterization at the sample level.

Methodology:

Generates multi-view representations by constructing 17 distinct feature types grouped into six categories and compresses cell-level data into interpretable sample-level summary statistics.