DiSTect

DiSTect implements a Bayesian shrinkage model for disease-associated gene discovery and prediction in spatial transcriptomics, incorporating spatial correlation among tissue spots through autoregressive terms.

Key Features:

• Spatial correlation modeling: Incorporates spatial correlation among tissue spots through autoregressive terms.
• Hierarchical structure: Facilitates the analysis of multiple correlated samples.
• Missing data accommodation: Extended to handle missing data within tissues.
• Dual computational frameworks: Two computational frameworks for Bayesian parameter estimation tailored to small and large sample scenarios.

Scientific Applications:

• HER2+ breast cancer analysis: Applied to analyze spatial transcriptomics data from studies of HER2+ breast cancer.
• Alzheimer's disease analysis: Applied to analyze spatial transcriptomics data from studies of Alzheimer's disease.

Methodology:

Bayesian shrinkage model with autoregressive terms for spatial correlation, hierarchical structure for multiple correlated samples, and two computational frameworks for Bayesian parameter estimation.