Binding affinity in drug design: experimental and computational techniques

ABSTRACT

Introduction: In pharmaceutical design where future drugs are developed by targeting a specific chosen protein, the evaluation of ligand affinity is crucial. For this very purpose are a multitude of diverse methods which are continuously being improved, which, in turn, makes it difficult to choose which techniques to use in practice.

Areas covered: In this review, the authors discuss both experimental and computational approaches for affinity evaluation. Basic principles, general limitations and advantages, as well as main areas of application in drug discovery, are overviewed for some of the most popular ligand binding assays. The authors further provide a guide to affinity predictions, collectively covering several techniques that are used in the first stages of rational drug design.

Expert opinion: All affinity estimation methods have limitations and advantages that partially overlap and complement one another. Some of the suggested best practices include cross-verification of data using at least two different techniques and careful data interpretation.

KEYWORDS:

• Drug design
• ligand
• inhibitor
• affinity
• free energy
• Gibbs energy
• binding
• ITC
• DSC
• DSF
• FTSA
• TSA
• AUC
• electrophoresis
• MST
• thermophoresis
• fluorescence
• ICD
• BLI
• SPR
• NMR
• QSAR
• molecular dynamics
• QM/MM
• MM-PBSA

Article highlights

• Affinity determination is one of the cornerstones of modern drug design.

• There are a lot of experimental and computational techniques available to evaluate a drug’s affinity towards its target, each with their own pros and cons.

• Isothermal titration calorimetry (ITC) measures the binding energy of unmodified interactors directly at physiological temperature but is costly both materials- and time-wise.

• High-throughput stability, mobility, or spectroscopic shift assays allow for rapid affinity evaluation, which is readily exploited for screening, but often are limited target-wise.

• In silico binding affinity prediction methods which encompass a vast range of algorithms are advanced enough to be able to reduce the Root Mean Square Error of prediction for a series of compounds to ~5 kJ/mol.

• Specialized competitions for calculating the binding affinity show the latest trends in the computational methods and their performance, compared to older, venerable approaches.

This box summarizes the key points contained in the article.

Acknowledgements

We apologise to all authors whose work we were unable to mention due to space constraints. M. Kazlauskiene is a European Molecular Biology Organization Fellow (ALTF 1087–2018).

Declaration of interest

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.