Ying-Chih Chiang

The Chinese University of Hong Kong (Shenzhen), China

Title: Optimising the ?-lactam Parameters Using the Force Field Toolkit

---

Biography

Ying-Chih Chiang graduated from National Taiwan University with a M. Sc. degree in Chemistry in 2006. After two years of work, she obtained a scholarship from International Max-Planck Research School for Quantum Dynamics and joined the Theoretical Chemistry Group in Heidelberg, Germany, under the supervision of Prof. L. S. Cederbaum. One year after obtaining her Ph. D. in 2012, she joined Prof. Y. Wang’s biophysics group in The Chinese University of Hong Kong. Since 2018 she joined Prof. J. W. Essex’s group in Southampton University as a Newton International Fellow. Starting from 2020 March she became an assistant professor at the Chinese University of Hong Kong – Shenzhen. Her research interests focus on using MD simulations and free energy calculations to reveal the resistance mechanism of antibiotics.

Abstract

Benefiting from the development of highly accurate force fields and the increase of computational power, molecular dynamics (MD) simulations are now frequently used in drug design. Often used all atom force fields include CHARMM, Amber, and OPLS-AA. However, while the associated ligand force fields such as CGenFF and GAFF are provided, accurate ligand force field parameters are not always readily available for drug molecules, and the users must optimise these parameters prior to the MD simulations. We examined the parameters of ?-lactam antibiotics, and discovered that the parameters of their core structures are labeled with high penalties by ParamChem, a web server that automatically generates CHARMM compatible parameters (CGenFF) for molecules. Therefore, a thorough optimization is needed. We performed this force field parameter optimisation for various ?-lactams using the Force Field Toolkit (FFTK) and Gaussian calculations. Two problems that users are likely to encounter when optimising other drug molecules are identified. First, inappropriate CGenFF parameters without penalty prediction could cause a difficulty in optimisation. Second, multiple dihedral parameter sets may produce the same molecular mechanics (MM) potentials of similar quality, raising the question on how to choose the right parameter set as the best solution. A systematic protocol incorporating molecular dynamics simulations and a principle for selecting the dihedral phase shifts are introduced. Using this protocol, we then successfully optimised both neutral and anionic forms of penam and of cephem. Our results highlight the importance of selecting proper phase shifts during the dihedral optimisation, and the protocol proposed in this work is beneficial to other users.