SST

SST assigns secondary structures to protein coordinate data using Minimum Message Length (MML) Bayesian inference to identify statistically supported secondary-structure elements.

Key Features:

• Bayesian MML inference: SST employs Minimum Message Length (MML) Bayesian inference, treating secondary-structure assignments as hypotheses and maximizing the joint probability of hypothesis and coordinate data.
• Rigorous hypothesis testing: SST incorporates a natural null hypothesis and rejects assignments that do not significantly improve upon this null model.
• Low-resolution performance: SST demonstrates reliable performance on low-resolution protein structures.
• Comparative evaluation: SST has been evaluated against DSSP and STRIDE in comparative studies.

Scientific Applications:

• Protein Structure Analysis: Identifying alpha-helices, beta-sheets, and other secondary-structure elements from protein coordinate data.
• Comparative Structural Studies: Providing consistent secondary-structure assignments for comparative analyses across proteins or experimental conditions.
• Structural Biology Research: Supporting investigations into the relationship between protein structure and function.

Methodology:

Each potential secondary-structure assignment is treated as a hypothesis tested against the observed coordinate data; hypotheses are optimized by maximizing their joint probability with the data via MML inference; the method has been compared and validated against DSSP and STRIDE.