HTM-ONE

HTM-ONE predicts one-dimensional equilibrium structural features (ESFs) of helical transmembrane regions from single amino acid sequences, including residue-level solvent accessibility, backbone dihedral angles, helix–helix contacts, and bend angles.

Key Features:

• Integrated Prediction Model: Simultaneously predicts multiple one-dimensional ESFs from single amino acid sequences using an integrated model, improving accuracy relative to separate models trained per feature.
• Improved Performance Metrics: Joint training of multiple ESFs yields better predictive outcomes than independent models, notably for solvent accessibility and bend angles.
• Bend-Angle Prediction: Predicts bend angles and identifies helical positions with severe kinks at a modest success rate.
• Conformational Dynamics Analysis: Estimates residue B-factors via normal mode analysis derived from observed and predicted ESFs to assess single-chain conformational dynamics.

Scientific Applications:

• Membrane Protein Structure–Function Studies: Enables analysis of residue-level structural properties relevant to membrane protein folding, stability, and interactions.
• Conformational Change Analysis: Identifies flexible residues and predicts local atomic fluctuations relevant to transport, ligand binding, and conformational transitions.
• Predictive Modeling of Structural Features: Provides a framework for developing integrated predictive models of one-dimensional structural properties in transmembrane helices.

Methodology:

Statistical analysis of common patterns among selected ESFs and joint training of multiple structural features, with estimation of B-factors via normal mode analysis based on observed and predicted ESFs.