SAHG

SAHG provides predicted structural models and functional annotations for human proteins, focusing on multi-domain proteins and intrinsically disordered regions to support analysis of protein structure–function relationships.

Key Features:

• Comprehensive Protein Structure Prediction: Employs alignment methods BLAST, PSI-BLAST, Smith–Waterman profile–profile alignment and the global-local alignment FORTE to analyze human protein sequences.
• Intrinsically Disordered Regions (IDR) Analysis: Incorporates POODLE-S for prediction of intrinsically disordered regions.
• Homology Modeling and Conformational Change Prediction: Uses MODELLER for homology modeling and eF-seek, elastic network models, and linear response theory to predict conformational changes, including ligand-induced shifts.
• Animated Structural Models: Provides predicted protein models as animated images to represent structural dynamics.
• Extensive Database Content: As of July 2010, contains 42,581 protein-domain models derived from approximately 24,900 unique human protein sequences sourced from the RefSeq database.
• Functional Annotation and Integration with Other Databases: Annotates models with functional information and links to EzCatDB, InterPro, and HPRD.

Scientific Applications:

• Structural analysis of multi-domain proteins: Supports analysis of domain architecture and structural modeling in multi-domain human proteins.
• Study of intrinsically disordered regions: Enables exploration of IDRs and their impact on protein structure and function.
• Protein-ligand interactions and conformational dynamics: Facilitates investigation of protein-ligand interactions and predicted conformational changes.
• Structure-function annotation and comparative analysis: Assists linking structural models to functional annotations via integration with EzCatDB, InterPro, and HPRD.
• Drug discovery and molecular biology research: Provides structural hypotheses useful for target characterization and related molecular biology studies.

Methodology:

SAHG's computational pipeline combines BLAST, PSI-BLAST, Smith–Waterman profile–profile alignment and FORTE for alignments; POODLE-S for IDR prediction; MODELLER for homology modeling; eF-seek, elastic network models, and linear response theory for conformational-change prediction; and uses sequences sourced from RefSeq.