Towards Predictive Coarse-grained Models

William Noid of the Pennsylvania State University is supported by an award from the Chemical Theory, Models and Computational Methods program in the Division of Chemistry to develop theory and computational methods for improving the predictive power of coarse-grained models in the chemical and materials sciences. Atomically detailed simulations provide exquisite insight into molecular structure, dynamics, and interactions. However, due to their computational cost, atomically detailed simulations can only effectively investigate very small length- and time-scales. In contrast, by eliminating unnecessary atomic details, coarse-grained (CG) model promise the necessary efficiency for simulating many processes of fundamental and technological significance that are far beyond the scope of atomically detailed models, e.g., the mechanisms by which viruses invade host cells or the phase behavior of industrially important polymers. Unfortunately, existing CG models provide a relatively poor description of thermodynamic properties. Moreover, CG models often demonstrate poor transferability, i.e., they require reparameterization for each system and environment of interest. These fundamental limitations severely curtail the predictive powers of current CG models. William Noid and his research group will derive, implement, and assess both theory and computational methods for ensuring that CG models are not only efficient, but also provide predictive accuracy and transferability for modeling soft materials, such as liquids and biomolecules. In addition, William Noid will continue developing an intergenerational science club that engages students of all ages in scientific discourse and discovery.

William Noid and his research group will develop rigorous theory and robust computational methods for addressing fundamental limitations of bottom-up CG models. Noid and his research group will analyze the many-body potential of mean force (PMF) to reveal fundamental insight and derive practical approaches for improving both the transferability and the thermodynamic properties of bottom-up models. The resulting insight will inform a dual approach for addressing the density-dependence of CG pair potentials, as well as the temperature-and composition-dependence of many-body local density potentials. Noid and his research group will also investigate the dual approach for describing the thermodynamic driving forces for self-assembly with CG models. Noid and his research group will investigate the influence of the CG mapping upon the exact PMF and upon the properties of approximate CG models. Noid and his group will develop and distribute software for implementing these methods as part of the Bottom-up Open-source Coarse-graining Software (BOCS) package. Noid will provide mentorship and rigorous training for graduate students. Moreover, Noid and his group will develop an intergenerational science club that integrates local senior citizens, emeritus faculty, and undergraduate students in order to build bridges between the academic and civic communities, educate the public about contemporary scientific topics, share the joy of scientific discovery, and promote a healthy lifestyle of life-long learning.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.