Spectre

Spectre provides reproducible computational workflows for preprocessing, batch alignment, integration, clustering, dimensionality reduction, visualization, population labelling, and quantitative analysis of high-dimensional single-cell cytometry data to characterize cellular heterogeneity across experiments.

Key Features:

• End-to-end workflow integration: Implements raw data pre-processing, batch alignment, data integration, clustering, dimensionality reduction, visualization, population labelling, and quantitative and statistical analyses.
• Batch and experiment integration: Integrates data across batches and experiments to enable cross-sample comparisons and combined analyses.
• Scalable analysis with machine learning: Supports application of machine learning classifiers and scalable data structures for very large datasets.
• Versatility across technologies: Applies to flow cytometry, mass cytometry (CyTOF), spectral cytometry, single-cell RNA sequencing (scRNAseq), and Imaging Mass Cytometry (IMC).
• Open and flexible data structures: Uses flexible data representations to accommodate diverse cytometry and single-cell data formats.
• Modular workflows: Provides modular analysis components that can be combined to build customized computational workflows.

Scientific Applications:

• Immunology: Enables identification and quantification of immune cell populations and states from high-dimensional cytometry data.
• Oncology: Facilitates profiling of tumor and microenvironment cellular heterogeneity using cytometry and imaging mass cytometry datasets.
• Developmental biology: Supports analysis of cell population dynamics and differentiation trajectories across developmental stages.
• Single-cell data integration and exploration: Allows integrated exploration of single-cell datasets to characterize cellular heterogeneity and underlying biological mechanisms.

Methodology:

Computational steps explicitly include raw data pre-processing, batch alignment, data integration, clustering, dimensionality reduction, visualization, population labelling, quantitative and statistical analyses, and application of machine learning classifiers using open, flexible data structures and modular workflows.