Fitmunk

Fitmunk fits side-chain conformations onto a fixed protein backbone within electron density maps to improve crystallographic model building, refinement, and validation.

Key Features:

• Automatic side-chain modeling: Uses deterministic dead-end elimination (DEE) theory to predict optimal side-chain positions by sampling dense conformer libraries.
• Rotamer prediction: Employs rotamer libraries informed by experimental data and rotamer-frequency priors to guide conformation selection.
• Hybrid energy evaluation: Integrates electron-density fit and prior rotamer information via a hybrid energy function to score side-chain conformations.
• Utilization of electron-density maps: Directly fits side chains into electron density maps to leverage experimentally observed density features for refinement and validation.
• Extensive conformational sampling: Performs dense conformational sampling to enhance accuracy of side-chain placement, particularly at medium and low resolution.
• Crystallographic applications and performance: Applies to initial model building, iterative refinement, and validation, accurately modeling over 95% of side chains across more than 115 new structures and a subset of 1100 Protein Data Bank (PDB) entries.
• Model validation: Assesses whether existing side chains are optimally fitted into electron density to inform further refinement.

Scientific Applications:

• Structural biology: Improves atomic-detail macromolecular models to support interpretation of protein structure, function, and interactions.
• Model refinement and validation: Provides targeted validation and refinement of crystallographic models, especially for proteins with complex side-chain arrangements.
• Low- and medium-resolution structure determination: Enhances side-chain placement accuracy in structures that are challenging to resolve using conventional methods.

Methodology:

Deterministic dead-end elimination (DEE) applied to dense conformer libraries; rotamer prediction using rotamer libraries informed by experimental data; hybrid energy function combining electron-density fit and rotamer priors; extensive conformational sampling.