pySuStaIn

SuStaIn: Data-driven disease subtyping and staging framework

SuStaIn infers latent disease subtypes and progression stages from cross-sectional data using probabilistic modeling of event sequences.

Key Features:

• Data-Driven Subtyping and Staging: Identifies distinct subtypes within heterogeneous patient populations and estimates individual disease stages.
• Flexible Modeling Approaches: Implements event-based models, piecewise linear z-score models, and scored events models to characterize subtype-specific progression patterns.
• Cross-Sectional Data Inference: Reconstructs temporal progression trajectories from cross-sectional datasets without requiring longitudinal data.

Scientific Applications:

• Progressive Disorder Research: Characterizes spatiotemporal disease progression patterns and stratifies patient subgroups for clinical research and therapeutic development.

Methodology:

Applies statistical inference to model ordered sequences of disease-related events or biomarker abnormalities. Estimates subtype-specific progression trajectories and probabilistic stage assignments at both population and individual levels.