AWSEM-Suite

AWSEM-Suite predicts monomeric protein tertiary structures from amino acid sequences using a coarse-grained protein force field and energy landscape theory to integrate physical and knowledge-based restraints.

Key Features:

• Template-guided modeling: Employs homologous template information to guide predictions and improve outcomes when templates are available.
• Coevolutionary restraints: Incorporates coevolutionary data to refine restraints and funnel the energy landscape toward native-like structures while maintaining flexibility.
• Coarse-grained protein force field: Implements an advanced coarse-grained protein force field rooted in neural network concepts.
• Physically motivated potentials: Integrates physically motivated potentials with other energetic terms to capture relevant interactions.
• Knowledge-based local structure biasing terms: Uses knowledge-based local structure biasing terms to favor realistic local conformations.
• Energy landscape theory optimization: Optimizes the balance of physical and bioinformational energies using energy landscape theory.
• CASP13 performance: Validated in CASP13, ranking eighth in both free-modeling and hard-to-model categories.

Scientific Applications:

• Protein structure prediction: Predicts 3D structures of proteins from sequence, including targets lacking readily available templates.
• Functional exploration: Enables functional studies via molecular dynamics and importance sampling simulations grounded in statistical mechanics.

Methodology:

Combines template-based modeling with coevolutionary restraints within a coarse-grained protein force field rooted in neural network concepts that integrates physically motivated potentials and knowledge-based local structure biasing terms and is optimized using energy landscape theory.