Evaluating the free‐energy landscape of proteins and the corresponding functional aspects presents a major challenge for computer simulation approaches. This challenge is due to the complexity of the landscape and the enormous computer time needed for converging simulations. The use of simplified coarse‐grained (CG) folding models offers an effective way of sampling the landscape but such a treatment, however, may not give the correct description of the effect of the actual protein residues. A general way around this problem that has been put forward in our early work (Fan et al., Theor Chem Acc 1999;103:77–80) uses the CG model as a reference potential for free‐energy calculations of different properties of the explicit model. This method is refined and extended here, focusing on improving the electrostatic treatment and on demonstrating key applications. These applications include: evaluation of changes of folding energy upon mutations, calculations of transition‐states binding free energies (which are crucial for rational enzyme design), evaluations of catalytic landscape, and evaluations of the time‐dependent responses to pH changes. Furthermore, the general potential of our approach in overcoming major challenges in studies of structure function correlation in proteins is discussed. Proteins 2010. © 2009 Wiley‐Liss, Inc.

In this review we give an overview of the field of Computational enzymology. We start by describing the birth of the field, with emphasis on the work of the 2013 chemistry Nobel Laureates. We then present key features of the state-of-the-art in the field, showing what theory, accompanied by experiments, has taught us so far about enzymes. We also briefly describe computational methods, such as quantum mechanics-molecular mechanics approaches, reaction coordinate treatment, and free energy simulation approaches. We finalize by discussing open questions and challenges.

Zinc metalloenzymes play a major role in key biological processes and Carboxypeptidase-A (CPA) is a major prototype of such enzymes. The present work quantifies the energetics of the catalytic reaction of CPA and its mutants using the EVB approach. The simulations allow us to quantify the origin of the catalytic power of this enzyme and to examine different mechanistic alternatives. The first step of the analysis used experimental information to determine the activation energy of each assumed mechanism of the reference reaction without the enzyme. The next step of the analysis involved EVB simulations of the reference reaction and then a calibration of the simulations by forcing them to reproduce the energetics of the reference reaction, in each assumed mechanism. The calibrated EVB was then used in systematic simulations of the catalytic reaction in the protein environment, without changing any parameter. The simulations reproduced the observed rate enhancement in two feasible general acid-general base mechanisms (GAGB-1 and GAGB-2), although the calculations with the GAGB-2 mechanism underestimated the catalytic effect in some treatments. We also reproduced the catalytic effect in the R127A mutant. The mutation calculations indicate that the GAGB-2 mechanism is significantly less likely than the GAGB-1 mechanism. It is also found, that the enzyme loses all its catalytic effect without the metal. This and earlier studies show that the catalytic effect of the metal is not some constant electrostatic effect, that can be assessed from gas phase studies, but a reflection of the dielectric effect of the specific environment.

Background: As not all target proteins can be easily screened in vitro, advanced virtual screening is becoming critical. Objective: In this study we demonstrate the application of reinforcement learning guided virtual screening for γaminobutyric acid A receptor (GABAAR) modulating peptides. Method: Structure-based virtual screening was performed on a receptor homology model. Screened molecules deemed to be novel were synthesized and analyzed using patch-clamp analysis. Results: 13 molecules were synthesized and 11 showed positive allosteric modulation, with two showing 50% activation at the low micromolar range. Conclusion: Reinforcement learning guided virtual screening is a viable method for the discovery of novel molecules that modulate a difficult to screen transmembrane receptor.