Protein–ligand interaction databases: advanced tools to mine activity data and interactions on a structural level

Abstract

The formation of molecular complexes between proteins and small organic substances is a fundamental concept of life. Biochemical experiments from X‐ray crystallography to isothermal titration calorimetry (ITC) are applied in large‐scale providing data for the analysis of the structural foundations of binding affinity. In recent years, several, mostly publically available databases emerged containing affinity data and structural information. These databases are central for the construction of complex models describing interaction geometries and correlate structural features to the strength of binding. Binding affinity databases reflect the knowledge of affinity measurements from many sources, mostly scientific and patent literature. A critical aspect is the data quality, which is affected by transcription errors during database construction as well as experimental uncertainties. The Protein Data Bank (PDB) is the central resource for macromolecular biological structures containing nearly 100,000 data entries today. Sophisticated geometric databases have been constructed based on this allowing for complex queries about the spatial arrangement of functional groups and their interactions. For scientists working in molecular design like medicinal chemists, access to this information can substantially support the process of creating new molecular entities specifically interacting with proteins of interest. WIREs Comput Mol Sci 2014, 4:562–575. doi: 10.1002/wcms.1192 This article is categorized under: Structure and Mechanism > Molecular Structures Computer and Information Science > Chemoinformatics Computer and Information Science > Databases and Expert Systems

Thermodynamic, kinetic, and structural parameterization of human carbonic anhydrase interactions toward enhanced inhibitor design

“…All these databases (Inhester and Rarey, 2014) have been of great importance but since most data were presented without a significant selection of intrinsic data, it is difficult to observe systematic correlations that would help to understand the protein–ligand recognition energetics. Figure 1 shows a general illustration of the intrinsic thermodynamic parameters.…”

Artificial intelligence in the prediction of protein–ligand interactions: recent advances and future directions

“…Here, for ML-based approaches, they mostly concentrate on the supervised and semisupervised models. A similar kind of analysis was carried out by Inhester and Rarey [ 40 ] describing the publicly available databases containing the affinity data and structural information that plays a vital role in describing interaction geometries and strength of binding.…”

Observing non-covalent interactions in experimental electron density for macromolecular systems: A novel perspective for protein–ligand interaction research

“…Noncovalent interactions (NCIs) govern protein-ligand interactions and are critical for understanding the determinants affecting ligand-binding affinity. To achieve a deep understanding of NCIs, many proteinligand interaction databases have been established in the last decade [2][3][4][5][6][7][8] . Two types of technologies are mainly applied to build such databases: 1.…”

“…To achieve a deep understanding of NCIs, many protein-ligand interaction databases have been established in the last decade. [1][2][3][4][5][6][7] Two types of technologies are primarily applied to build such databases: (i) structure-based data mining and (ii) quantum mechanical (QM) methods-powered computation. For the first type, protein-ligand complex structures in the Protein Data Bank (PDB) are used as the main source, and different indices, such as distance, angle, exposed surface, and line-of-sight statistics, are used to depict the possibility of NCIs between a pair or two groups of atoms.…”