Berlin 2018 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

CPP: Fachverband Chemische Physik und Polymerphysik

CPP 20: Poster Session I

CPP 20.26: Poster

Montag, 12. März 2018, 17:30–19:30, Poster A

Quantifying how coarse-graining reduces the size of chemical compound space — •Kiran Kanekal, Kurt Kremer, and Tristan Bereau — Max Planck Institute for Polymer Research

Increasing the efficiency of materials design and discovery remains a significant challenge, especially given the prohibitively large size of chemical compound space. Efficient sampling of chemical compound space can be achieved in silico with the use of transferable coarse-grained (CG) models that retain the essential properties of a higher resolution method. In addition to reducing computational expense, use of a chemically transferable CG model enables different molecular fragments to map to the same bead type. This further increases sampling efficiency, effectively reducing the size of chemical compound space. For example, the MARTINI[1] force field consists of 14 different neutral bead types, allowing for 119 unique representations consisting of 1-bead and 2-bead CG molecules. We previously showed that over 400,000 molecules could be mapped to these 119 MARTINI representations, demonstrating a drastic reduction of chemical compound space[2]. However, it is unclear as to how much variability exists within the subset of molecules and functional groups that map to a single bead type. In this work, we investigate these subsets for each MARTINI bead type to quantify the effective range of chemical space that is covered by that bead type. We further propose new criteria for the rational design of CG models that allows for the optimization of their chemical transferability. [1] Fink and Reymond, J. Chem. Inf. Model. 2007, 47, 342. [2] Menichetti et al., J. Chem. Phys. 2017, 147, 125101.