Reciprocal best structure hits

Reciprocal best structure hits identifies homologous protein pairs by comparing AlphaFold 2.0 predicted structures to detect distant homologues that sequence-based methods may miss.

Key Features:

• Structure-Based Homology Detection: Uses AlphaFold 2.0 predicted protein structure models to identify homologous proteins based on structural similarity rather than sequence alignment.
• Comprehensive Database Utilization: Analyzes structure models from the AlphaFold database covering five model organisms to provide a diverse dataset for homology detection.
• Comparison with Sequence-Based Methods: Compares structural matching results with traditional sequence similarity approaches to assess overlap and identify homologues missed by sequence analysis.
• Novel Homologue Discovery: Enables discovery of potential novel homologues solely through structural matching, extending detection beyond sequence divergence.
• Limitations in Disordered or Short Proteins: Acknowledges reduced performance when handling highly disordered proteins or short protein structures.

Scientific Applications:

• Functional genomics: Identifies homologous relationships that can aid functional annotation of genes and proteins.
• Evolutionary biology: Detects distant homologues to inform studies of protein evolution and divergence.
• Protein interaction studies: Reveals structural homologs that can inform hypotheses about protein interactions and complexes.
• Annotation of uncharacterized proteins: Provides structural evidence to support characterization of uncharacterized proteins and elucidation of biological pathways and mechanisms.

Methodology:

Uses AlphaFold 2.0 predicted structure models from the AlphaFold database for five model organisms, performs structural comparisons to identify reciprocal best structure hits, and compares those structural matches to results from traditional sequence similarity approaches.