procoil

procoil predicts whether a coiled-coil sequence, defined by its amino acid composition and heptad register, will form a dimer or a trimer to determine protein oligomeric state.

Key Features:

• Machine learning-based prediction: Employs models that recognize weighted amino acid patterns governing coiled-coil oligomerization to classify sequences as dimeric or trimeric.
• Prediction profiles: Computes residue-level scores that quantify each position's contribution to dimer versus trimer propensity.
• Visualization of profiles: Produces curve and heatmap representations of residue contributions to illustrate local oligomeric tendencies.

Scientific Applications:

• Oligomerization rule elucidation: Analyzes sequence determinants that govern formation of two- and three-stranded coiled coils.
• Protein structure–function studies: Assesses how coiled-coil oligomeric state influences protein structural and functional relationships.
• Case study — GCN4: Characterizes the yeast transcription factor GCN4 as a hybrid oligomer with both dimeric and trimeric characteristics, consistent with theoretical and experimental data.

Methodology:

Develops machine learning models trained on sequence datasets to identify weighted amino acid patterns associated with dimer versus trimer formation and generates residue-level contribution visualizations.