Scellnetor

Scellnetor applies network-constrained time-series clustering to scRNA-seq trajectories to identify gene modules with differential temporal expression within molecular interaction networks.

Key Features:

• Network-constrained time-series clustering: Scellnetor implements a network-constrained time-series clustering algorithm that clusters genes based on connectivity in molecular interaction networks to extract systems-biology signatures from scRNA-seq data.
• Pseudo-temporal trajectory comparison: It compares two sets of cells or developmental trajectories to identify differentially expressed or similarly expressed gene modules driving cellular differentiation.
• Scanpy AnnData (H5AD) support: Scellnetor processes scRNA-seq data provided as Scanpy-generated AnnData objects in H5AD file format.
• Connected gene component output: The tool outputs connected gene components (subnetworks) reflecting differential or similar temporal expression patterns within molecular interaction networks.
• Expression visualization and statistics: Scellnetor provides visualizations of mean gene expression with confidence intervals and tables of statistically significant Gene Ontology (GO) terms.
• Systems biology interpretation: By integrating network connectivity with temporal expression, Scellnetor yields mechanistically interpretable gene modules for systems-level hypotheses.

Scientific Applications:

• Hematopoiesis in mice: Applied to study the molecular control of hematopoiesis, identifying key gene networks that drive this process.
• CD8 T-cell development: Used to explore dysfunctional CD8 T-cell development in chronic infections, providing mechanistically interpretable subnetworks relevant to immune dysregulation.

Methodology:

Processes Scanpy-generated AnnData objects in H5AD format; applies network-constrained time-series clustering to compare two cell sets or trajectories and output connected gene components based on differential or similar expression patterns; generates mean gene expression visualizations with confidence intervals and tables of statistically significant Gene Ontology (GO) terms.