Exploring differential evolution for inverse QSAR analysis

Abstract

Inverse quantitative structure-activity relationship (QSAR) modeling encompasses the generation of compound structures from values of descriptors corresponding to high activity predicted with a given QSAR model. Structure generation proceeds from descriptor coordinates optimized for activity prediction. Herein, we concentrate on the first phase of the inverse QSAR process and introduce a new methodology for coordinate optimization, termed differential evolution (DE), that originated from computer science and engineering. Using simulation and compound activity data, we demonstrate that DE in combination with support vector regression (SVR) yields effective and robust predictions of optimized coordinates satisfying model constraints and requirements. For different compound activity classes, optimized coordinates are obtained that exclusively map to regions of high activity in feature space, represent novel positions for structure generation, and are chemically meaningful.

Learning Molecular Representations for Medicinal Chemistry

“…Whereas standard QSAR models map molecular structures to an activity or property, inverse QSAR models turn this notion on its head, instead seeking to generate new molecular structures that satisfy optimal properties or activities . On paper, this approach is highly attractive as a hypothesis generator for new molecules with desirable properties, yet historically, inverse-design models have been unable to map a continuous activity or property back to discrete, viable molecules. , The intractability of this problem has instead led researchers to adopt virtual screening approaches to rapidly evaluate pre-enumerated compound libraries of synthetically accessible molecules …”

“…24 On paper, this approach is highly attractive as a hypothesis generator for new molecules with desirable properties, yet historically, inverse-design models have been unable to map a continuous activity or property back to discrete, viable molecules. 115,116 The intractability of this problem has instead led researchers to adopt virtual screening approaches to rapidly evaluate pre-enumerated compound libraries of synthetically accessible molecules. 117 Generative models in deep learning now directly address the inverse design problem, leading to new opportunities for de novo drug design.…”

Quantum Chemistry Calculations for Metabolomics

“…Recent advances in deep learning − have shown promise in predicting physiochemical properties previously reliant on quantum chemical calculations, such as CCS, NMR chemical shifts, and MS fragmentation patterns, ,− as well as replacing quantum chemical calculations entirely. − Deep learning has demonstrated improvements in property prediction accuracy compared to quantum-chemistry-based approaches, while reducing per-structure computation time by orders of magnitude (hours for quantum chemical, versus milliseconds with deep learning after training). , Similar improvements are seen with deep-learning-based DFT: orders of magnitude reduction in computation time with sub-1% MAE . For generative approaches, including autoencoder and adversarial networks, − deep learning offers additional potential in addressing the inverse quantitative structure–property relationship (QSPR) problem, − wherein molecular structure candidates can be determined from physiochemical property/properties. However, with deep learning follows considerations to training time, which better contextualizes per-structure computation time, and generalizability, or to what extent does prediction accuracy degrade as inputs vary from the training set.…”