Dataset Information

The pK(a) Distribution of Drugs: Application to Drug Discovery.

ABSTRACT: The acid-base dissociation constant (pK(a)) of a drug is a key physicochemical parameter influencing many biopharmaceutical characteristics. While this has been well established, the overall proportion of non-ionizable and ionizable compounds for drug-like substances is not well known. Even less well known is the overall distribution of acid and base pK(a) values. The current study has reviewed the literature with regard to both the proportion of ionizable substances and pK(a) distributions. Further to this a set of 582 drugs with associated pK(a) data was thoroughly examined to provide a representative set of observations. This was further enhanced by delineating the compounds into CNS and non-CNS drugs to investigate where differences exist. Interestingly, the distribution of pK(a) values for single acids differed remarkably between CNS and non-CNS substances with only one CNS compound having an acid pK(a) below 6.1. The distribution of basic substances in the CNS set also showed a marked cut off with no compounds having a pK(a) above 10.5.The pK(a) distributions of drugs are influenced by two main drivers. The first is related to the nature and frequency of occurrence of the functional groups that are commonly observed in pharmaceuticals and the typical range of pK(a) values they span. The other factor concerns the biological targets these compounds are designed to hit. For example, many CNS targets are based on seven transmembrane G protein-coupled receptors (7TM GPCR) which have a key aspartic acid residue known to interact with most ligands. As a consequence, amines are mostly present in the ligands that target 7TM GPCR's and this influences the pK(a) profile of drugs containing basic groups. For larger screening collections of compounds, synthetic chemistry and the working practices of the chemists themselves can influence the proportion of ionizable compounds and consequent pK(a) distributions. The findings from this study expand on current wisdom in pK(a) research and have implications for discovery research with regard to the composition of corporate databases and collections of screening compounds. Rough guidelines have been suggested for the profile of compound collections and will evolve as this research area is expanded.

SUBMITTER: Manallack DT

PROVIDER: S-EPMC2754920 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundTechnological and research advances have produced large volumes of biomedical data. When represented as a network (graph), these data become useful for modeling entities and interactions in biological and similar complex systems. In the field of network biology and network medicine, there is a particular interest in predicting results from drug-drug, drug-disease, and protein-protein interactions to advance the speed of drug discovery. Existing data and modern computational methods allow to identify potentially beneficial and harmful interactions, and therefore, narrow drug trials ahead of actual clinical trials. Such automated data-driven investigation relies on machine learning techniques. However, traditional machine learning approaches require extensive preprocessing of the data that makes them impractical for large datasets. This study presents wide range of machine learning methods for predicting outcomes from biomedical interactions and evaluates the performance of the traditional methods with more recent network-based approaches.ResultsWe applied a wide range of 32 different network-based machine learning models to five commonly available biomedical datasets, and evaluated their performance based on three important evaluations metrics namely AUROC, AUPR, and F1-score. We achieved this by converting link prediction problem as binary classification problem. In order to achieve this we have considered the existing links as positive example and randomly sampled negative examples from non-existant set. After experimental evaluation we found that Prone, ACT and [Formula: see text] are the top 3 best performers on all five datasets.ConclusionsThis work presents a comparative evaluation of network-based machine learning algorithms for predicting network links, with applications in the prediction of drug-target and drug-drug interactions, and applied well known network-based machine learning methods. Our work is helpful in guiding researchers in the appropriate selection of machine learning methods for pharmaceutical tasks.

Dataset Information

The pK(a) Distribution of Drugs: Application to Drug Discovery.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets