Ocelli

Ocelli implements diffusion-based modeling for scalable visualization and exploration of developmental multimodal single-cell data, focusing on cell state transitions and biological processes involving developmental progression.

Key Features:

• Low-dimensional embedding visualization: Ocelli visualizes cells on a low-dimensional embedding that preserves the continuity of developmental progression.
• Rare and transient state identification: Ocelli identifies rare and transient cell states in developmental data.
• Trajectory inference integration: Ocelli integrates with trajectory inference algorithms for developmental analysis.
• Feature count imputation: Ocelli imputes undetected feature counts in single-cell data.
• Scalable implementation: Ocelli is implemented in Ray for scalable computation on multimodal single-cell data.

Scientific Applications:

• Developmental multimodal single-cell analysis: Ocelli analyzes simultaneous genome-wide measurements of multiple modalities including gene expression, chromatin accessibility, protein epitopes, and multiple histone modifications at single-cell resolution.
• Cell trajectory visualization: Ocelli visualizes cell trajectories to study developmental processes at multiple layers of gene regulation.
• Cell state transition modeling: Ocelli models biological processes involving cell state transitions during development.

Methodology:

Ocelli uses diffusion-based modeling as its core computational approach for analyzing developmental single-cell data.