NNcon

NNcon predicts protein residue contact maps using 2D-recursive neural networks to support protein folding rate analysis, model selection, and three-dimensional structure prediction.

Key Features:

• Contact map prediction: Predicts residue–residue contacts from protein sequence information.
• 2D-recursive neural networks: Implements 2D-recursive neural networks as the core prediction algorithm.
• Accuracy (CASP8): Ranked among the most accurate contact predictors in the Eighth Critical Assessment of Techniques for Protein Structure Prediction (CASP8, 2008).
• Speed: Designed for fast execution to accommodate large-scale analyses.

Scientific Applications:

• Protein folding rate prediction: Provides contact information that informs analyses of protein folding kinetics.
• Model selection: Enables evaluation and selection of structural models based on predicted contacts.
• Three-dimensional structure prediction: Supplies contact restraints useful for reconstructing protein 3D structures.

Methodology:

NNcon employs 2D-recursive neural networks to analyze protein sequences and predict residue contact maps.