DSP

DSP predicts protein shape strings to represent protein conformations at higher resolution than traditional secondary structure models, including regions classified as random coils.

Key Features:

• Shape string representation: Provides protein shape strings as a structural descriptor that encapsulates comprehensive conformational information, including random-coil regions.
• Knowledge-driven sequence alignment: Employs a knowledge-driven sequence alignment approach for alignment-based inference of shape features.
• Sequence shape string profile method: Implements a sequence shape string profile method to model sequence-to-shape relationships.
• Prediction outputs: Generates predicted shape strings for query sequences and produces sequence–shape string profiles.
• Higher-resolution modeling: Offers a representation more detailed than traditional secondary structure models.
• Validation: Performance has been validated using blind test data.

Scientific Applications:

• Comparative structural analysis: Enables comparative analysis of protein structures within shape string space.
• Evolutionary studies: Supports evolutionary studies by enabling comparison of proteins based on shape string profiles.
• Evolutionary trajectory visualization: Facilitates visualization of evolutionary trajectories derived from shape string data.

Methodology:

Uses knowledge-driven sequence alignment and a sequence shape string profile method to predict protein shape strings, with performance evaluated on blind test data.