TR&D Project 4. The Modeling Stage: Multiscale Spatiotemporal Modeling of Macromolecular Systems in Cellular Neighborhoods Here, we are concerned with the modeling of (i) static structures of large macromolecular machines and (ii) dynamic processes involving these machines. The accuracy, precision, and completeness of these models can in principle be improved by using all available information (ie, integrative modeling), including from various experiments, physical theories, statistical analyses, and prior models. We will improve several aspects of integrative modeling, tuning these improvements to the biological systems and data described in TR&Ds, DBPs, and C&SPs, and then implementing them in our open-source Integrative Modeling Platform (IMP) package. To improve integrative modeling of structures of macromolecular complexes, we will: (i) develop integrative modeling relying on near-atomic resolution cryo-electron microscopy density maps; (ii) develop explicit atomic- resolution representation of chemical cross-links; (iii) develop a scoring function for integrative modeling of related complexes from multiple species; and (iv) develop a sensitive Bayesian scoring function for assessing a model against cryo-EM single particle images. To improve integrative modeling of dynamics of macromolecular systems, we will: (i) estimate kinetic rate constants for changes in structure and composition based on time- dependent cryo-EM images; and (ii) develop integrative modeling of long macromolecular processes by satisfaction of spatiotemporal data. Finally, we will also leverage our method development efforts by supporting other developers and users of IMP.
Showing the most recent 10 out of 111 publications
