Structural conservation in single-domain proteins: implications for homology modeling

Area of Science:

• Structural Bioinformatics
• Computational Biology
• Protein Structure Prediction

Background:

• Large-scale sequencing projects generate vast amounts of protein data, outpacing experimental structure determination.
• A significant gap exists between known protein sequences and experimentally determined structures.
• Homology (comparative) modeling offers a computational approach to infer protein structures.

Purpose of the Study:

• To assess the reliability of sequence alignments for homology modeling.
• To evaluate the accuracy of homology models concerning structural conservation.
• To investigate the conservation patterns in insertion and deletion regions for improved loop modeling.

Main Methods:

• Analysis of sequence alignments as a basis for homology modeling.
• Evaluation of structural conservation in homologous protein pairs.
• Comparative analysis of core regions versus insertion/deletion regions in protein structures.

Main Results:

• Sequence information effectively guides the construction of acceptable protein models, particularly for the structural core.
• Regions with insertions and deletions exhibit significantly lower structural conservation compared to the protein core.
• The findings highlight limitations in predicting structural variations in loop regions using sequence data alone.

Conclusions:

• Homology modeling is a valuable tool for predicting protein structures and narrowing the sequence-structure gap.
• Sequence-based predictions are most reliable for conserved protein cores.
• Further advancements are needed for accurate loop modeling, considering the lower conservation in these regions.