virtual screening

In this work, we combined Deep Docking and free energy MD simulations for the in silico screening and experimental validation for potential inhibitors of leucine rich repeat kinase 2 (LRRK2) targeting the WD40 repeat (WDR) domain.

Traditional computational approaches to design chemical species are limited by the need to compute properties for a vast number of candidates, e.g., by discriminative modeling. Therefore, inverse design methods aim to start from the desired property and optimize a corresponding chemical structure. From a machine learning viewpoint, the inverse design problem can be addressed through so-called generative modeling. Mathematically, discriminative models are defined by learning the probability distribution function of properties given the molecular or material structure. In contrast, a generative model seeks to exploit the joint probability of a chemical species with target characteristics. The overarching idea of generative modeling is to implement a system that produces novel compounds that are expected to have a desired set of chemical features, effectively sidestepping issues found in the forward design process. In this contribution, we overview and critically analyze popular generative algorithms like generative adversarial networks, variational autoencoders, flow, and diffusion models. We highlight key differences between each of the models, provide insights into recent success stories, and discuss outstanding challenges for realizing generative modeling discovered solutions in chemical applications.

Overactivation of the N-methyl-D-aspartate receptor (NMDAR) in postsynaptic neurons leads to glutamate-related excitotoxicity in the central nervous system of mammals. We have built 3-D models of each domain for the universal screening of potential toxicants and their binding mechanisms. Our docking results show that the calculated pK (i) values of glycine and L: -glutamate significantly increase (>1) when the NR1 and NR2A S1S2 domains are closing, respectively. Inversely, D: -cycloserine (DCS) and 5,7-dichlorokynurenic acid (5,7-DCKA) do not show such a dependence on domain closure. Replica exchange molecular dynamics (REMD) confirmed 5 different conformational states of the S1S2 domain along the 308.2 K temperature trajectory. Analysis of residue fluctuations during this temperature trajectory showed that residues in loop 1, loop 2, the amino terminal domain (ATD), and the area linked to ion channel α-helices are involved in this movement. This further implicates the notion that efficacious ligands act through S1S2 lobe movement which can culminate in the opening or closing of the ion channel. We further tested this by docking hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) and octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocane (HMX) to the S1S2 domain. Our results predict that these nitramines are not efficacious and thus do not produce excitoxicity when they bind to the S1S2 domain of the NMDAR.