SEGMER

SEGMER identifies conserved protein substructure motifs by segmental threading of target sequences against the Protein Data Bank (PDB) to improve structural modeling and functional inference.

Key Features:

• Segmental Threading Approach: SEGMER divides target sequences into segments of two to four consecutive or nonconsecutive secondary structural elements and threads these segments through the PDB to detect matching substructure motifs.
• Integration with Whole-Chain Threading: SEGMER integrates segmental results with whole-chain threading to enhance overall alignment and model quality.
• Quantified Accuracy Improvements: Integration with whole-chain threading yields a 16% increase in TM-score and a 25% improvement in spatial restraint accuracy, with a 28% TM-score increase and 48% contact accuracy improvement observed on 12 CASP8 free-modeling targets.
• Validation on Hard Proteins: The method was tested on 144 nonredundant hard proteins to evaluate performance on challenging cases.

Scientific Applications:

• Protein structure modeling: Improves precision of structural predictions and alignments in comparative and de novo modeling.
• Functional inference: Facilitates inference of protein function by identifying conserved substructure motifs.
• Drug discovery: Supports target characterization and structural interpretation relevant to drug discovery.
• Enzyme design: Assists identification of conserved motifs useful for enzyme engineering and design.
• Molecular biology: Aids studies of protein interactions and mechanisms by refining structural predictions.

Methodology:

Segment target sequences into units based on secondary structure; thread these segments through the PDB to find corresponding motifs; integrate segmental matches with whole-chain threading results to enhance prediction accuracy.