DCS

DCS performs automated analysis of single-cell RNA sequencing (scRNA-seq) data to identify cell types, quantify anomalous cells, and visualize cell phenotypic landscapes.

Key Features:

• Automatic Cell Type Identification: Provides a voting algorithm that aggregates multiple classification results and a Hopfield classifier that leverages energy-like functions to assign cell types from gene expression profiles.
• Cell Anomaly Quantification: Uses an isolation forest approach to detect and quantify anomalous cells within large scRNA-seq datasets.
• Visualization of Cell Phenotypic Landscapes: Generates visualizations based on Hopfield energy-like functions to explore and interpret cellular phenotypes.
• Quality Control: Implements quality control procedures for scRNA-seq datasets.
• Batch Correction: Performs batch effect correction for scRNA-seq data.
• Clustering: Provides clustering methods to group cells by expression profiles.

Scientific Applications:

• PBMC analysis: Analysis of peripheral blood mononuclear cells (PBMC) scRNA-seq datasets.
• Bone marrow plasma cell analysis: Analysis of plasma cells from bone marrow, including samples from healthy donors and patients with multiple myeloma.
• Deconvolution of cell mixtures: Deconvolving heterogeneous cell mixtures in complex scRNA-seq samples.
• Anomalous cell detection: Detecting and quantifying small numbers of anomalous cells within large cohorts.

Methodology:

Implemented in Python and comprising a voting algorithm, a Hopfield classifier using energy-like functions for classification and visualization, and an isolation forest for anomaly detection.