PREDDIMER

PREDDIMER predicts transmembrane helix dimer structures from input TM protein fragment sequences by modeling fragments as ideal α-helices and ranking predicted conformations.

Key Features:

• Dimer Structure Reconstruction: Reconstructs multiple dimer structures from sequences of TM protein fragments modeled as ideal α-helices.
• Scoring and Filtering: Applies a scoring function to rank and filter predicted dimer conformations for prioritization.
• 3D Dimer Structure Visualization: Produces three-dimensional models of predicted TM helix dimers to inspect spatial arrangements and intermolecular contacts.
• Surface Hydrophobicity Maps: Generates 2D maps of surface hydrophobicity and interface regions of TM helices to assess interface properties.
• Ensemble Prediction: Outputs an ensemble of possible dimer conformations to represent conformational variability of bitopic membrane proteins.

Scientific Applications:

• Bitopic membrane protein dimerization studies: Supports structural analysis of dimerization in bitopic membrane proteins.
• Receptor tyrosine kinase research: Facilitates investigation of TM helix interactions relevant to receptor tyrosine kinases.
• Drug design and protein engineering: Provides structural hypotheses that can inform drug design and protein engineering efforts.
• Structural biology and hypothesis generation: Enables generation of testable structural models for experimental validation.

Methodology:

Benchmarking used 11 TM sequences with experimentally determined 3D dimer conformations from nuclear magnetic resonance (NMR) spectroscopy; predicted backbone RMSDs are typically below 3 Å, and a randomization test reported a strong anticorrelation (−0.82) between scoring function values and the statistical significance of chance-based predictions.