Characterisation of data resources for in silico modelling: benchmark datasets for ADME properties

Abstract

The cost of in vivo and in vitro screening of ADME properties of compounds has motivated efforts to develop a range of in silico models. At the heart of the development of any computational model are the data; high quality data are essential for developing robust and accurate models. The characteristics of a dataset, such as its availability, size, format and type of chemical identifiers used, influence the modelability of the data. Areas covered: This review explores the usefulness of publicly available ADME datasets for researchers to use in the development of predictive models. More than 140 ADME datasets were collated from publicly available resources and the modelability of 31 selected datasets were assessed using specific criteria derived in this study. Expert opinion: Publicly available datasets differ significantly in information content and presentation. From a modelling perspective, datasets should be of adequate size, available in a user-friendly format with all chemical structures associated with one or more chemical identifiers suitable for automated processing (e.g. CAS number, SMILES string or InChIKey). Recommendations for assessing dataset suitability for modelling and publishing data in an appropriate format are discussed.

Assessment and Reproducibility of Quantitative Structure–Activity Relationship Models by the Nonexpert

“…Identifying QSAR models for ADME endpoints available in the literature Models available in the literature for ADME endpoints were identified using information provided in relevant reviews [19][20][21] in addition to searching on-line literature databases (PubMed, Google). The search terms that were used were taken from a previously identified list of ADME parameters, 24 which are provided here as Supporting Information. Where multiple publications were identified for a particular ADME parameter, a sample of up to six papers was selected for further analysis; in these instances more recent publications (e.g.…”

“…For the first principle, 'a model must have a defined endpoint', a methodology was considered to have fulfilled this principle fully if it clearly described the modelling of one of the previously enumerated ADME parameters. 24 For the second principle, 'a model must have an unambiguous algorithm', if a clear description of both the QSAR modelling procedure and the final equation were provided, then this principle was considered to have been met in full.…”

Advances in the prediction of gastrointestinal absorption: Quantitative Structure-Activity Relationship (QSAR) modelling of PAMPA permeability

“…It comprises Double-Sink PAMPA intrinsic and effective permeability coefficients determined for nearly 300 compounds (mostly commercial drugs) and was noted a benchmark dataset for PAMPA by Przybylak et al [12]. Most of the larger datasets consist of commercial drugs as well and are constructed to evaluate PAMPA assay modifications.…”

“…In addition to the attractive methodology, PAMPA is a promising source of reliable experimental data for the computational (in silico) evaluation of a drug's GI absorption [12]. Such models, typically Quantitative Structure-Activity Relationships (QSARs), attempt to relate PAMPA permeability to physico-chemical properties and structural descriptors.…”

A critical review of adverse effects to the kidney: mechanisms, data sources, andin silicotools to assist prediction

“…, physico-chemical and toxicological data are the bedrock of the development of in silico models for toxicology. Some of the general issues related to data procurement for modelling purposes have been discussed elsewhere[108][109][110][111][112][113]. If the development of AOPs and multi-scale models is currently considered to be the panacea of 21 st century toxicology, data spanning molecular to population levels are necessary to generate multi-scale models resembling the structure of AOPs.For the registration of many chemicals and pharmaceuticals, adverse effects to the kidney and bladder are currently assessed through traditional toxicological approaches, involving in vitro and in vivo animal studies[114].…”