Design of a general-purpose European compound screening library for EU-OPENSCREEN.

This work describes a collaborative effort to define and apply a protocol for the rational selection of a general-purpose screening library, to be used by the screening platforms affiliated with the EU-OPENSCREEN initiative. It is designed as a standard source of compounds for primary screening against novel biological targets, at the request of research partners. Given the general nature of the potential applications of this compound collection, the focus of the selection strategy lies on ensuring chemical stability, absence of reactive compounds, screening-compliant physicochemical properties, loose compliance to drug-likeness criteria (as drug design is a major, but not exclusive application), and maximal diversity/coverage of chemical space, aimed at providing hits for a wide spectrum of drugable targets. Finally, practical availability/cost issues cannot be avoided. The main goal of this publication is to inform potential future users of this library about its conception, sources, and characteristics. The outline of the selection procedure, notably of the filtering rules designed by a large committee of European medicinal chemists and chemoinformaticians, may be of general methodological interest for the screening/medicinal chemistry community. The selection task of 200K molecules out of a pre-filtered set of 1.4M candidates was shared by five independent European research groups, each picking a subset of 40K compounds according to their own in-house methodology and expertise. An in-depth analysis of chemical space coverage of the library serves not only to characterize the collection, but also to compare the various chemoinformatics-driven selection procedures of maximal diversity sets. Compound selections contributed by various participating groups were mapped onto general-purpose self-organizing maps (SOMs) built on the basis of marketed drugs and bioactive reference molecules. In this way, the occupancy of chemical space by the EU-OPENSCREEN library could be directly compared with distributions of known bioactives of various classes. This mapping highlights the relevance of the selection and shows how the consensus reached by merging the five different 40K selections contributes to achieve this relevance. The approach also allows one to readily identify subsets of target- or target-class-oriented compounds from the EU-OPENSCREEN library to suit the needs of the diverse range of potential users. The final EU-OPENSCREEN library, assembled by merging five independent selections of 40K compounds from various expert groups, represents an excellent example of a Europe-wide collaborative effort toward the common objective of building best-in-class European open screening platforms.

Applicability domains for classification problems: Benchmarking of distance to models for Ames mutagenicity set.

The estimation of accuracy and applicability of QSAR and QSPR models for biological and physicochemical properties represents a critical problem. The developed parameter of "distance to model" (DM) is defined as a metric of similarity between the training and test set compounds that have been subjected to QSAR/QSPR modeling. In our previous work, we demonstrated the utility and optimal performance of DM metrics that have been based on the standard deviation within an ensemble of QSAR models. The current study applies such analysis to 30 QSAR models for the Ames mutagenicity data set that were previously reported within the 2009 QSAR challenge. We demonstrate that the DMs based on an ensemble (consensus) model provide systematically better performance than other DMs. The presented approach identifies 30-60% of compounds having an accuracy of prediction similar to the interlaboratory accuracy of the Ames test, which is estimated to be 90%. Thus, the in silico predictions can be used to halve the cost of experimental measurements by providing a similar prediction accuracy. The developed model has been made publicly available at http://ochem.eu/models/1 .

Stochastic versus stepwise strategies for quantitative structure-activity relationship generation--how much effort may the mining for successful QSAR models take?

Descriptor selection in QSAR typically relies on a set of upfront working hypotheses in order to boil down the initial descriptor set to a tractable size. Stepwise regression, computationally cheap and therefore widely used in spite of its potential caveats, is most aggressive in reducing the effectively explored problem space by adopting a greedy variable pick strategy. This work explores an antipodal approach, incarnated by an original Genetic Algorithm (GA)-based Stochastic QSAR Sampler (SQS) that favors unbiased model search over computational cost. Independent of a priori descriptor filtering and, most important, not limited to linear models only, it was benchmarked against the ISIDA Stepwise Regression (SR) tool. SQS was run under various premises, varying the training/validation set splitting scheme, the nonlinearity policy, and the used descriptors. With the considered three anti-HIV compound sets, repeated SQS runs generate sometimes poorly overlapping but nevertheless equally well validating model sets. Enabling SQS to apply nonlinear descriptor transformations increases the problem space: nevertheless, nonlinear models tend to be more robust validators. Model validation benchmarking showed SQS to match the performance of SR or outperform it in cases when the upfront simplifications of SR "backfire", even though the robust SR got trapped in local minima only once in six cases. Consensus models from large SQS model sets validate well--but not outstandingly better than SR consensus equations. SQS is thus a robust QSAR building tool according to standard validation tests against external sets of compounds (of same families as used for training), but many of its benefits/drawbacks may yet not be revealed by such tests. SQS results are a challenge to the traditional way to interpret and exploit QSAR: how to deal with thousands of well validating models, nonetheless providing potentially diverging applicability ranges and predicted values for external compounds. SR does not impose such burden on the user, but is "betting" on a single equation or a narrow consensus model to behave properly in virtual screening a sound strategy? By posing these questions, this article will hopefully act as an incentive for the long-haul studies needed to get them answered.