nan Consonni

Comparison of different approaches to define the applicability domain of QSAR models.

One of the OECD principles for model validation requires defining the Applicability Domain (AD) for the QSAR models. This is important since the reliable predictions are generally limited to query chemicals structurally similar to the training compounds used to build the model. Therefore, characterization of interpolation space is significant in defining the AD and in this study some existing descriptor-based approaches performing this task are discussed and compared by implementing them on existing validated datasets from the literature. Algorithms adopted by different approaches allow defining the interpolation space in several ways, while defined thresholds contribute significantly to the extrapolations. For each dataset and approach implemented for this study, the comparison analysis was carried out by considering the model statistics and relative position of test set with respect to the training space.

Chemometrics in QSAR

Quantitative Structure–Activity Relationships (QSARs) and Quantitative Structure–Property Relationships (QSPRs) are scientific fields in which the use of chemometric methods is of outstanding importance. Indeed, chemometric methods, as well as statistics and chemoinformatics, are the basic tools for finding mathematical meaningful relationships between the molecular structure and biological activities, physicochemical, toxicological, and environmental properties of chemicals. Three main topics are involved in the QSAR/QSPR approach to the scientific research: the concept of molecular structure, the definition of molecular descriptors, and the chemometric tools. After a short historical presentation, QSAR modeling approaches, both those based on the classical chemometric methods and those defined by some peculiar strategies, are presented. Then, after an introduction to molecular descriptors, some basic modeling methodologies based on the variable descriptor selection are described. Finally, the general principles for QSAR/QSPR analysis are reported, paying special attention to the concept of applicability domain (AD) and model validation. These principles have been recently received from the Organization for Economic Co-operation and Development (OECD) as the principles for accepting QSAR models for regulatory purposes and adopted as guidelines for the new European Community Regulation on Chemicals (Registration, Evaluation and Authorisation of Chemicals (REACH)).

A similarity-based QSAR model for predicting acute toxicity towards the fathead minnow (Pimephales promelas)

REACH regulation demands information about acute toxicity of chemicals towards fish and supports the use of QSAR models, provided compliance with OECD principles. Existing models present some drawbacks that may limit their regulatory application. In this study, a dataset of 908 chemicals was used to develop a QSAR model to predict the LC50 96 hours for the fathead minnow. Genetic algorithms combined with k nearest neighbour method were applied on the training set (726 chemicals) and resulted in a model based on six molecular descriptors. An automated assessment of the applicability domain (AD) was carried out by comparing the average distance of each molecule from the nearest neighbours with a fixed threshold. The model had good and balanced performance in internal and external validation (182 test molecules), at the expense of a percentage of molecules outside the AD. Principal Component Analysis showed apparent correlations between model descriptors and toxicity.

Investigating the mechanisms of bioconcentration through QSAR classification trees

This paper proposes a scheme to predict whether a compound (1) is mainly stored within lipid tissues, (2) has additional storage sites (e.g., proteins), or (3) is metabolized/eliminated with a reduced bioconcentration. The approach is based on two validated QSAR (Quantitative Structure-Activity Relationship) trees, whose salient features are: (a) descriptor interpretability and (b) simplicity. Trees were developed for 779 organic compounds, the TGD approach was used to quantify the lipid-driven bioconcentration, and a refined machine-learning optimization procedure was applied. We focused on molecular descriptor interpretation, which allowed us to gather new mechanistic insights into the bioconcentration mechanisms.

Validation and extension of a similarity-based approach for prediction of acute aquatic toxicity towards Daphnia magna

Quantitative structure–activity relationship (QSAR) models for predicting acute toxicity to Daphnia magna are often associated with poor performances, urging the need for improvement to meet REACH requirements. The aim of this study was to evaluate the accuracy, stability and reliability of a previously published QSAR model by means of further external validation and to optimize its performance by means of extension to new data as well as a consensus approach. The previously published model was validated with a large set of new molecules and then compared with ChemProp model, from which most of the validation data were taken. Results showed better performance of the proposed model in terms of accuracy and percentage of molecules outside the applicability domain. The model was re-calibrated on all the available data to confirm the efficacy of the similarity-based approach. The extended dataset was also used to develop a novel model based on the same similarity approach but using binary fingerprints to describe the chemical structures. The fingerprint-based model gave lower regression statistics, but also less unpredicted compounds. Eventually, consensus modelling was successfully used to enhance the accuracy of the predictions and to halve the percentage of molecules outside the applicability domain.

Self Organizing Maps For Analysis Of Polycyclic Aromatic Hydrocarbons 3-Way Data From Spilled Oils

In this paper, the application of a new method based on self-organizing maps (SOM; termed MOLMAP, molecular map of atom-level properties) to handle 3-way data generated in a monitoring environmental study is presented. The study comprised 50 polycyclic aromatic hydrocarbons (PAHs) analyzed in samples derived from the weathering of six oil products (four crude oils and two fuel oils) spilled under controlled conditions for about 4 months. MOLMAP yielded useful information on each mode of the data cube: weathering samples, spilled oil products, and PAHs. Thus, the different behaviors of the six oils were ascertained, along with their particular evolution on time, and their weathering patterns were studied in terms of the original PAHs. Thus, the two heaviest products (two fuel oils) were characterized by two neurons whose more relevant weights were associated to heavy PAHs, as C(1)-fluoranthene, C(2)-fluoranthene, benzo(ghi)perylene, and dibenz(ah)anthracene. The six spilled products were projected on different regions on both the MOLMAP-SOM and a subsequent principal components analysis (PCA) scatter plot, developed using the so-called MOLMAP-scores. Besides, it was possible to further differentiate between unweathered, or slightly weathered, samples and the most weathered ones. The more relevant PAHs characterizing those samples were assessed studying the weights of the neurons in which the samples got projected.

Mapping of Activity through Dichotomic Scores (MADS): A new chemoinformatic approach to detect activity-rich structural regions

A new chemoinformatic approach, called Mapping of Activity through Dichotomic Scores, is introduced. Its goal is the supervised projection of molecules, represented with strings of binary digits expressing the presence or absence of selected structural features, onto a novel 2‐dimensional space, which highlights regions of active (inactive) molecules of interest. At the same time, variables are projected onto a second 2‐dimensional space, which highlights those structural features that are more related to the molecular activity of interest. Unlike the classical weighting schemes used in substructural analysis, which consider the substructures independently of each other, the Mapping of Activity through Dichotomic Scores approach considers the interactions between pairs of substructures, that is, their frequencies of cooccurrence in the molecules. In this work, the theory is presented and elucidated, with an example dataset and in comparison with a benchmark fragment‐based scoring scheme.