Cyclum

Cyclum identifies and characterizes circular trajectories in single-cell RNA sequencing (scRNA-seq) gene expression data to enable cell cycle analysis and removal of cell cycle-related confounding.

Key Features:

• AutoEncoder for circular trajectories: Employs an AutoEncoder architecture specifically tailored to identify and characterize circular trajectories in high-dimensional gene expression spaces.
• scRNA-seq focus: Operates on single-cell RNA sequencing (scRNA-seq) data for analysis of periodic biological processes such as the cell cycle.
• Circular trajectory mapping: Maps circular trajectories in expression space to recover detailed cell cycle information.
• Confounder removal: Enables removal of cell cycle-related factors from expression data to reduce confounding effects on downstream analyses.
• Accuracy and robustness: Improves the accuracy and robustness of cell cycle characterization relative to existing methodologies.

Scientific Applications:

• Cell cycle characterization: Dissects cell cycle phase and progression in scRNA-seq datasets.
• Subpopulation delineation: Enhances delineation of cell subpopulations within complex tissues and tumor environments by mitigating cell cycle variability.
• Cell atlas construction: Supports construction of comprehensive cell atlases through refined cell state identification.
• Tumor heterogeneity analysis: Facilitates investigation of tumor heterogeneity by separating cell cycle effects from other sources of variation.
• Developmental and disease studies: Applied to developmental processes and disease mechanisms that involve periodic biological events like the cell cycle.

Methodology:

Uses an AutoEncoder approach adapted to identify and characterize circular trajectories in high-dimensional gene expression space derived from scRNA-seq data.