Predicting protein flexibility and stability

点击翻译按钮获取中文摘要

A grand challenge of biophysics is to understand protein folding, stability, flexibility, and function in terms of structure and solvent condition. A novel Distance Constraint Model (DCM) is employed to accurately predict protein stability in aqueous solution under specified thermodynamic conditions (i.e. temperature, pH, ionic strength, etc) from known three-dimensional structure. This project builds upon prior success of the PI in developing efficient rigidity-graph algorithms to identify flexible and rigid regions in proteins modeled as a fixed constraint topology, and development of the DCM. The DCM is based on the hypothesis that network rigidity is an underlying mechanism for enthalpy-entropy compensation, yielding a mathematically precise algorithm to account for non-additivity in free energy decompositions. A proof of concept, minimal DCM, will be extended in this project to include explictit modeling of essential entropy-compensation mechanisms that include (a) hydration, (b) hydrophobic interactions, (c) electrostatics interactions, with (d) a residue-specific parameterization. These extensions will allow prediction of protein stability in mixed solvent conditions, and bring the DCM closer to a fully transferable parameterization. However, parameter transferability is not a requirement of this proposed work, as the utility of our minimal DCM has been firmly established. The first outcome.of this work will be the release of a fast computational tool that harmoniously quantifies stabilitiy and flexibility in practical computing times necessary for protien design applications. For example, local- details of protein flexibility are quantified to identify correlated atomic motions important for induced fit of ligand binding and allosteric conformational changes. Synergistic application of the DCM with protein family evolutionary descriptions will provide key insight into familial variability of Quantified Stability/Flexibility Relationships (QSFR). The second outcome will be a public accessible QSFR database providing users wide access to DCM results and analysis tools will give users a practical means to better understand protein function in realistic computing times needed for the post-geonomic era.