There are two key requirements for reliably simulating enzyme reactions: one is a reasonably accurate potential energy surface to describe the bond forming/breaking process as well as to adequately model the heterogeneous enzyme environment; the other is to perform extensive sampling since an enzyme system consists of at least thousands of atoms and its energy landscape is very complex. One attractive approach to meet both daunting tasks is Born-Oppenheimer ab initio QM/MM molecular dynamics simulation (aiQM/MM-MD) with umbrella sampling. In this chapter, we describe our recently developed pseudobond Q-Chem–Amber interface, which employs a combined electrostatic-mechanical embedding scheme with periodic boundary condition and the particle mesh Ewald method for long-range electrostatics interactions. In our implementation, Q-Chem and the sander module of Amber are combined at the source code level without using system calls, and all necessary data communications between QM and MM calculations are achieved via computer memory. We demonstrate the applicability of this pseudobond Q-Chem–Amber interface by presenting two examples, one reaction in aqueous solution and one enzyme reaction. Finally, we describe our established aiQM/MM-MD enzyme simulation protocol, which has been successfully applied to study more than a dozen enzymes.

No abstract

ANOVA) of the degree and distribution of 116 morphological assessments from 10 tissues from 10 rats from 4 TRTS are nearing completion; preliminarily no toxic TRT effects exist.CONCLUSIONS: Biologically significant increases in any of 4 manifestations of toxicity (death, structural abnormalities, growth alterations, and functional deficits) are indicative of an agent's toxicity. This is the first preclinical evidence showing TRT with rTIMP1+Ab that restores ovulatory function in endometriosis does not cause toxicity, indicative of a safe translation of TIMP1 neutralizing therapies to restore ovarian function in women with endometriosis.