Investigating conformational transitions of proteins by coarse-grained elastic network models
MetadataShow full item record
Proteins are large molecular machines. Many of these machines carry out conformational transitions to perform function. It is very difficult to determine all metastable protein conformations experimentally. Therefore, computational methods have been developed to investigate metastable protein conformations and conformational transitions. For most of proteins, atomistic molecular dynamics cannot reach the time scales of conformational transitions, which are typically beyond microseconds. The large size of proteins is another obstacle in atomistic molecular dynamics simulations. Coarse-grained elastic network models can provide an alternative to overcome the time scale and size problems. In this dissertation, we have investigated conformational transitions of proteins by using modified elastic network models. These models can be applied in two ways. First, they allow us to analyze conformational transition pathways and deduce the dynamic order of structural events. Second, they enable us to build models for unknown protein conformations by incorporating experimental data. For the first application, a transition pathway modeling method called iENM will be presented in Chapter 2. For the second application, a flexible fitting method based on small angle X-ray scattering (SAXS) data will be discussed in Chapter 3. Our methods will be compared to alternative methods and they will be validated by experimental data.