RFCoil

RFCoil predicts oligomeric states of coiled-coil regions in proteins, distinguishing parallel dimeric and trimeric formations.

Key Features:

• Sequence Analysis: Uses non-redundant amino acid indices and a Random Forest classifier to predict oligomeric states from sequence.
• Structural Modeling: Analyzes sequence-to-structure relationships in coiled-coil folding and oligomer formation.
• High Accuracy: Achieves AUC of 0.849 in cross-validation and 0.855 on an independent test set, outperforming LOGICOIL, PrOCoil, SCORER 2.0, and Multicoil2.
• Mechanistic Insights: Extracts rules from the trained Random Forest model to elucidate determinants of oligomeric formation.

Scientific Applications:

• Protein Interaction Studies: Predicts coiled-coil oligomerization states relevant to materials science, synthetic biology, and medicine.

Methodology:

Combines non-redundant amino acid indices with Random Forest to model sequence-to-structure relationships in coiled-coil folding and oligomer formation.