Methods Applied for the Allosteric Site Revelation

Abstract

Allostery is the functional change at one site on a protein due to alterations at a distant site. Allosteric modulators by targeting allosteric site can fine‐tune protein's activity. Drugging allosteric sites has therefore been regarded as a novel direction in drug discovery. However, identification of allosteric sites by experimental approaches is serendipitous. Since bioinformatics and structural biology develop rapidly, protein dynamics and structures can be better understood and allosteric drug discovery is on the cusp of a novel era of rational computational methodologies. Various computational allosteric site prediction models have therefore been put forward. Computational prediction is helpful for exploration of potential allosteric sites due to its higher accuracy and relatively shorter timescale. Here, we review the state‐of‐the‐art computational allosteric sites prediction methods. We first classify different approaches according to their underlying theoretical models, then algorithms and fundamental theories behind them are revealed, and finally successful examples are presented. We put forward the current challenges faced by these approaches. Possible solutions are provided with the latest advances in the field of computational biology, and tactics to resolve these demanding tasks are also discussed.

Along the allostery stream: Recent advances in computational methods for allosteric drug discovery

“…Mostly uncovered by serendipity through exhaustive and tedious experiments, such as alanine mutagenesis screening, the recalcitrant allosteric sites have been gradually tamed through recent advances in computational methods, thereby promoting structure-based allosteric drug design. 28,34,36,41,155 Allosteric sites are classified as visible or cryptic ("hidden") sites, based on their structural characteristics within the protein conformational ensembles. The former is readily observed within protein crystal structures and is therefore easy to detect, while the latter only transiently emerges throughout protein conformational dynamics and has thus been historically intractable by most experimental techniques.…”

“…As a prerequisite for rational allosteric drug design, predicting allosteric pocket is of central importance and is regarded as the “holy grail” in allosteric studies, and has therefore received the utmost interest from both academia and pharmaceutical industry. Mostly uncovered by serendipity through exhaustive and tedious experiments, such as alanine mutagenesis screening, the recalcitrant allosteric sites have been gradually tamed through recent advances in computational methods, thereby promoting structure‐based allosteric drug design 28,34,36,41,155 …”