Advanced In Silico Approaches for Drug Discovery: Mining Information from Multiple Biological and Chemical Data Through mtk- QSBER and pt-QSPR Strategies

Abstract

The last decade has been seeing an increase of public-private partnerships in drug discovery, mostly driven by factors such as the decline in productivity, the high costs, time, and resources needed, along with the requirements of regulatory agencies. In this context, traditional computer-aided drug discovery techniques have been playing an important role, enabling the identification of new molecular entities at early stages. However, recent advances in chemoinformatics and systems pharmacology, alongside with a growing body of high quality, publicly accessible medicinal chemistry data, have led to the emergence of novel in silico approaches. These novel approaches are able to integrate a vast amount of multiple chemical and biological data into a single modeling equation. The present review analyzes two main kinds of such cutting-edge in silico approaches. In the first subsection, we discuss the updates on multitasking models for quantitative structure-biological effect relationships (mtk- QSBER), whose applications have been significantly increasing in the past years. In the second subsection, we provide detailed information regarding a novel approach that combines perturbation theory with quantitative structure-property relationships modeling tools (pt- QSPR). Finally, and most importantly, we show that the joint use of mtk-QSBER and pt- QSPR modeling tools are apt to guide drug discovery through its multiple stages: from in vitro assays to preclinical studies and clinical trials.

Keywords: Box-Jenkins moving averages; CHEMBL; knowledge generator; mtk-QSBER models; perturbation theory; pt-QSPR models.