HOMA

HOMA performs homology modeling by satisfaction of spatial restraints to predict three-dimensional protein structures from template proteins.

Key Features:

• Homology modeling algorithm: Implements homology modeling by satisfaction of spatial restraints (Li et al., Protein Sci 1997) to derive target structures from templates.
• Target–template alignment: Uses sequence alignment between target and template proteins to guide model building based on template coordinates.
• Model generation: Produces homology models from the target–template alignment and provides accompanying validation reports.
• Accuracy assessment: Evaluates model accuracy by superimposition against experimentally determined structures.
• Benchmarking results: Validated on 24 groups of homologous proteins yielding 510 models, of which 264 had correct folds and 246 had incorrect folds.
• Structure quality validation: Employs Verify3D and ProsaII and reports metrics such as steric conformational energy (including Van der Waals energy), MolProbity clashscore, and PROCHECK G-factors to distinguish correct versus incorrect folds and assess model accuracy.
• Comparative performance: Prediction accuracies are reported as comparable to other state-of-the-art automated homology modeling approaches.

Scientific Applications:

• Structure prediction: Predicts protein three-dimensional structures when experimental structural data are unavailable or incomplete.
• Functional inference: Supports inference of protein function from predicted structural features.
• Interaction and evolutionary studies: Facilitates analysis of protein–protein interactions and evolutionary relationships via modeled structures.
• Drug discovery and enzyme design: Provides structural models to inform drug discovery and enzyme engineering efforts.
• Disease mechanism investigation: Aids exploration of molecular disease mechanisms through structural hypotheses.

Methodology:

Align target sequences to template coordinates, generate models by satisfaction of spatial restraints, and validate models by superimposition against experimental structures and by structure quality tools including Verify3D, ProsaII, steric conformational energy (including Van der Waals), MolProbity clashscore, and PROCHECK G-factors.