Evaluation of the predictivity of Acute Oral Toxicity (AOT) structure-activity relationship models.

Several public efforts are aimed at discovering patterns or classifiers in the high-dimensional bioactivity space that predict tissue, organ or whole animal toxicological endpoints. The current study sought to assess and compare the predictions of the Globally Harmonized System (GHS) categories and Dangerous Goods (DG) classifications based on Lethal Dose (LD50) from several available tools (ACD/Labs, Leadscope, T.E.S.T., CATMoS, CaseUltra). External validation was done using dataset of 375 substances to demonstrate their predictive capacity. All models showed very good performance for identifying non-toxic compounds, which would be useful for DG classification, developing or triaging new chemicals, prioritizing existing chemicals for more detailed and rigorous toxicity assessments, and assessing non-active pharmaceutical intermediates. This would ultimately reduce animal use and improve risk assessments. Category-to-category prediction was not optimal, mainly due to the tendency to overpredict the outcome and the general limitations of acute oral toxicity (AOT) in vivo studies. Overprediction does not specifically pose a risk to human health, it can impact transport and material packaging requirements. Performance for compounds with LD50 ≤ 300 mg/kg (approx. 5% of the dataset) was the poorest among all groups and could be potentially improved by including expert review and read-across to similar substances.

Multi-Task Neural Networks and Molecular Fingerprints to Enhance Compound Identification from LC-MS/MS Data.

Mass spectrometry (MS) is widely used for the identification of chemical compounds by matching the experimentally acquired mass spectrum against a database of reference spectra. However, this approach suffers from a limited coverage of the existing databases causing a failure in the identification of a compound not present in the database. Among the computational approaches for mining metabolite structures based on MS data, one option is to predict molecular fingerprints from the mass spectra by means of chemometric strategies and then use them to screen compound libraries. This can be carried out by calibrating multi-task artificial neural networks from large datasets of mass spectra, used as inputs, and molecular fingerprints as outputs. In this study, we prepared a large LC-MS/MS dataset from an on-line open repository. These data were used to train and evaluate deep-learning-based approaches to predict molecular fingerprints and retrieve the structure of unknown compounds from their LC-MS/MS spectra. Effects of data sparseness and the impact of different strategies of data curing and dimensionality reduction on the output accuracy have been evaluated. Moreover, extensive diagnostics have been carried out to evaluate modelling advantages and drawbacks as a function of the explored chemical space.

From the Streets to the Judicial Evidence: Determination of Traditional Illicit Substances in Drug Seizures by a Rapid and Sensitive UHPLC-MS/MS-Based Platform.

According to the 2021 World Drug Report, around 275 million people use drugs of abuse, and 36 million people suffer from addiction, fostering a thriving market for illicit substances. In Italy, 30,083 people were reported to the Judicial Authority for offenses in violation of the Italian Law D.P.R. 309/1990. These offences are sentenced after a qualitative-quantitative analysis of seized materials. Given the large quantity of seized drugs and the need to perform accurate analytical determinations, Italian forensic laboratories struggle to complete analyses in a short time, delaying the entire reporting process needed to achieve sentencing. For this purpose, an UHPLC-MS/MS-based platform was developed at the University of Milano-Bicocca to support law-enforcement authorities. Software was designed to easily manage street seizure acquisition, documentation registration, and sampling. A sensitive UHPLC-MS/MS method was fully validated for the quantification of the traditional illicit substances (cocaine, heroin, 6-MAM, morphine, amphetamine, methamphetamine, MDMA, ketamine, GHB, GBL, LSD, trans-∆9-THC, and THCA) at the ppb level. The final report is relayed to the Prefecture in 3-4 days, even within 24 h for urgent requests. The platform allows for semi-automatic data handling to minimize erroneous results for an accurate report generation by standardized procedures.

Parsimonious Optimization of Multitask Neural Network Hyperparameters.

Neural networks are rapidly gaining popularity in chemical modeling and Quantitative Structure-Activity Relationship (QSAR) thanks to their ability to handle multitask problems. However, outcomes of neural networks depend on the tuning of several hyperparameters, whose small variations can often strongly affect their performance. Hence, optimization is a fundamental step in training neural networks although, in many cases, it can be very expensive from a computational point of view. In this study, we compared four of the most widely used approaches for tuning hyperparameters, namely, grid search, random search, tree-structured Parzen estimator, and genetic algorithms on three multitask QSAR datasets. We mainly focused on parsimonious optimization and thus not only on the performance of neural networks, but also the computational time that was taken into account. Furthermore, since the optimization approaches do not directly provide information about the influence of hyperparameters, we applied experimental design strategies to determine their effects on the neural network performance. We found that genetic algorithms, tree-structured Parzen estimator, and random search require on average 0.08% of the hours required by grid search; in addition, tree-structured Parzen estimator and genetic algorithms provide better results than random search.

ToF-SIMS and Machine Learning for Single-Pixel Molecular Discrimination of an Acrylate Polymer Microarray.

Combinatorial approaches to materials discovery offer promising potential for the rapid development of novel polymer systems. Polymer microarrays enable the high-throughput comparison of material physical and chemical properties-such as surface chemistry and properties like cell attachment or protein adsorption-in order to identify correlations that can progress materials development. A challenge for this approach is to accurately discriminate between highly similar polymer chemistries or identify heterogeneities within individual polymer spots. Time-of-flight secondary ion mass spectrometry (ToF-SIMS) offers unique potential in this regard, capable of describing the chemistry associated with the outermost layer of a sample with high spatial resolution and chemical sensitivity. However, this comes at the cost of generating large scale, complex hyperspectral imaging data sets. We have demonstrated previously that machine learning is a powerful tool for interpreting ToF-SIMS images, describing a method for color-tagging the output of a self-organizing map (SOM). This reduces the entire hyperspectral data set to a single reconstructed color similarity map, in which the spectral similarity between pixels is represented by color similarity in the map. Here, we apply the same methodology to a ToF-SIMS image of a printed polymer microarray for the first time. We report complete, single-pixel molecular discrimination of the 70 unique homopolymer spots on the array while also identifying intraspot heterogeneities thought to be related to intermixing of the polymer and the pHEMA coating. In this way, we show that the SOM can identify layers of similarity and clusters in the data, both with respect to polymer backbone structures and their individual side groups. Finally, we relate the output of the SOM analysis with fluorescence data from polymer-protein adsorption studies, highlighting how polymer performance can be visualized within the context of the global topology of the data set.

Self-Organizing Map and Relational Perspective Mapping for the Accurate Visualization of High-Dimensional Hyperspectral Data.

We present an optimization of the toroidal self-organizing map (SOM) algorithm for the accurate visualization of hyperspectral data. This represents a significant advancement on our previous work, in which we demonstrated the use of toroidal SOMs for the visualization of time-of-flight secondary ion mass spectrometry (ToF-SIMS) imaging data. We have previously shown that the toroidal SOM can be used, unsupervised, to produce a multicolor similarity map of the analysis area, in which pixels with similar mass spectra are assigned a similar color. Here, we use an additional algorithm, relational perspective mapping (RPM), to produce more accurate visualizations of hyperspectral data. The SOM output is used as an input for the RPM algorithm, which is a nonlinear dimensionality reduction technique designed to produce a two-dimensional map of high-dimensional data. Using the topological information provided by the SOM, RPM provides complementary distance information. The result is a color scheme that more accurately reflects the local spectral distances between pixels in the data. We exemplify SOM-RPM using ToF-SIMS imaging data from a mouse tumor tissue section. The similarity maps produced are compared with those produced by two leading hyperspectral visualization techniques in the field of mass spectrometry imaging: t-distributed stochastic neighborhood embedding (t-SNE) and uniform manifold approximation and projection (UMAP). We evaluate the performance of each technique both qualitatively and quantitatively, investigating the correlations between distances in the models and distances in the data. SOM-RPM is demonstrably highly competitive with t-SNE and UMAP, according to our evaluations. Furthermore, the use of a neural network offers distinct advantages in data characterization, which we discuss. We also show how spectra extracted from regions of interest identified by SOM-RPM can be further analyzed using linear discriminant analysis for the validation and characterization of the surface chemistry.

NURA: A curated dataset of nuclear receptor modulators.

Nuclear receptors (NRs) are key regulators of human health and constitute a relevant target for medicinal chemistry applications as well as for toxicological risk assessment. Several open databases dedicated to small molecules that modulate NRs exist; however, depending on their final aim (i.e., adverse effect assessment or drug design), these databases contain a different amount and type of annotated molecules, along with a different distribution of experimental bioactivity values. Stemming from these considerations, in this work we aim to provide a unified dataset, NURA (NUclear Receptor Activity) dataset, collecting curated information on small molecules that modulate NRs, to be intended for both pharmacological and toxicological applications. NURA contains bioactivity annotations for 15,247 molecules and 11 selected NRs, and it was obtained by integrating and curating data from toxicological and pharmacological databases (i.e., Tox21, ChEMBL, NR-DBIND and BindingDB). Our results show that NURA dataset is a useful tool to bridge the gap between toxicology- and medicinal-chemistry-related databases, as it is enriched in terms of number of molecules, structural diversity and covered atomic scaffolds compared to the single sources. To the best of our knowledge, NURA dataset is the most exhaustive collection of small molecules annotated for their modulation of the chosen nuclear receptors. NURA dataset is intended to support decision-making in pharmacology and toxicology, as well as to contribute to data-driven applications, such as machine learning. The dataset and the data curation pipeline can be downloaded free of charge on Zenodo at the following DOI: https://doi.org/10.5281/zenodo.3991561.

Consensus versus Individual QSARs in Classification: Comparison on a Large-Scale Case Study.

Consensus strategies have been widely applied in many different scientific fields, based on the assumption that the fusion of several sources of information increases the outcome reliability. Despite the widespread application of consensus approaches, their advantages in quantitative structure-activity relationship (QSAR) modeling have not been thoroughly evaluated, mainly due to the lack of appropriate large-scale data sets. In this study, we evaluated the advantages and drawbacks of consensus approaches compared to single classification QSAR models. To this end, we used a data set of three properties (androgen receptor binding, agonism, and antagonism) for approximately 4000 molecules with predictions performed by more than 20 QSAR models, made available in a large-scale collaborative project. The individual QSAR models were compared with two consensus approaches, majority voting and the Bayes consensus with discrete probability distributions, in both protective and nonprotective forms. Consensus strategies proved to be more accurate and to better cover the analyzed chemical space than individual QSARs on average, thus motivating their widespread application for property prediction. Scripts and data to reproduce the results of this study are available for download.

Machine Learning Consensus To Predict the Binding to the Androgen Receptor within the CoMPARA Project.

The nuclear androgen receptor (AR) is one of the most relevant biological targets of Endocrine Disrupting Chemicals (EDCs), which produce adverse effects by interfering with hormonal regulation and endocrine system functioning. This paper describes novel in silico models to identify organic AR modulators in the context of the Collaborative Modeling Project of Androgen Receptor Activity (CoMPARA), coordinated by the National Center of Computational Toxicology (U.S. Environmental Protection Agency). The collaborative project involved 35 international research groups to prioritize the experimental tests of approximatively 40k compounds, based on the predictions provided by each participant. In this paper, we describe our machine learning approach to predict the binding to AR, which is based on a consensus of a multivariate Bernoulli Naive Bayes, a Random Forest, and N-Nearest Neighbor classification models. The approach was developed in compliance with the Organization of Economic Cooperation and Development (OECD) principles, trained on 1687 ToxCast molecules classified according to 11 in vitro assays, and further validated on a set of 3,882 external compounds. The models provided robust and reliable predictions and were used to gather novel data-driven insights on the structural features related to AR binding, agonism, and antagonism.

Beware of Unreliable Q2! A Comparative Study of Regression Metrics for Predictivity Assessment of QSAR Models.

Validation is an essential step of QSAR modeling, and it can be performed by both internal validation techniques (e.g., cross-validation, bootstrap) or by an external set of test objects, that is, objects not used for model development and/or optimization. The evaluation of model predictive ability is then completed by comparing experimental and predicted values of test molecules. When dealing with quantitative QSAR models, validation results are generally expressed in terms of Q2 metrics. In this work, four fundamental mathematical principles, which should be respected by any Q2 metric, are introduced. Then, the behavior of five different metrics (QF12, QF22, QF32, QCCC2, and QRm2) is compared and critically discussed. The conclusions highlight that only the QF32 metric satisfies all the stated conditions, while the remaining metrics show different theoretical flaws.

N3 and BNN: Two New Similarity Based Classification Methods in Comparison with Other Classifiers.

Two novel classification methods, called N3 (N-nearest neighbors) and BNN (binned nearest neighbors), are proposed. Both methods are inspired by the principles of the K-nearest neighbors (KNN) method, being both based on object pairwise similarities. Their performance was evaluated in comparison with nine well-known classification methods. In order to obtain reliable statistics, several comparisons were performed using 32 different literature data sets, which differ for number of objects, variables and classes. Results highlighted that N3 on average behaves as the most efficient classification method with similar performance to support vector machine based on radial basis function kernel (SVM/RBF). The method BNN showed on average higher performance than the classical K-nearest neighbors method.

Prediction of acute aquatic toxicity toward Daphnia magna by using the GA-kNN method.

In this study, a QSAR model was developed from a data set consisting of 546 organic molecules, to predict acute aquatic toxicity toward Daphnia magna. A modified k-Nearest Neighbour (kNN) strategy was used as the regression method, which provided prediction only for those molecules with an average distance from the k nearest neighbours lower than a selected threshold. The final model showed good performance (R(2) and Q(2) cv equal to 0.78, Q(2) ext equal to 0.72). It comprised eight molecular descriptors that encoded information about lipophilicity, the formation of H-bonds, polar surface area, polarisability, nucleophilicity and electrophilicity.

Towards global QSAR model building for acute toxicity: Munro database case study.

A series of 436 Munro database chemicals were studied with respect to their corresponding experimental LD50 values to investigate the possibility of establishing a global QSAR model for acute toxicity. Dragon molecular descriptors were used for the QSAR model development and genetic algorithms were used to select descriptors better correlated with toxicity data. Toxic values were discretized in a qualitative class on the basis of the Globally Harmonized Scheme: the 436 chemicals were divided into 3 classes based on their experimental LD50 values: highly toxic, intermediate toxic and low to non-toxic. The k-nearest neighbor (k-NN) classification method was calibrated on 25 molecular descriptors and gave a non-error rate (NER) equal to 0.66 and 0.57 for internal and external prediction sets, respectively. Even if the classification performances are not optimal, the subsequent analysis of the selected descriptors and their relationship with toxicity levels constitute a step towards the development of a global QSAR model for acute toxicity.

Artificial Intelligence and Machine Learning Methods to Evaluate Cardiotoxicity following the Adverse Outcome Pathway Frameworks.

Cardiovascular disease is a leading global cause of mortality. The potential cardiotoxic effects of chemicals from different classes, such as environmental contaminants, pesticides, and drugs can significantly contribute to effects on health. The same chemical can induce cardiotoxicity in different ways, following various Adverse Outcome Pathways (AOPs). In addition, the potential synergistic effects between chemicals further complicate the issue. In silico methods have become essential for tackling the problem from different perspectives, reducing the need for traditional in vivo testing, and saving valuable resources in terms of time and money. Artificial intelligence (AI) and machine learning (ML) are among today's advanced approaches for evaluating chemical hazards. They can serve, for instance, as a first-tier component of Integrated Approaches to Testing and Assessment (IATA). This study employed ML and AI to assess interactions between chemicals and specific biological targets within the AOP networks for cardiotoxicity, starting with molecular initiating events (MIEs) and progressing through key events (KEs). We explored methods to encode chemical information in a suitable way for ML and AI. We started with commonly used approaches in Quantitative Structure-Activity Relationship (QSAR) methods, such as molecular descriptors and different types of fingerprint. We then increased the complexity of encoders, incorporating graph-based methods, auto-encoders, and character embeddings employed in neural language processing. We also developed a multimodal neural network architecture, capable of considering the complementary nature of different chemical representations simultaneously. The potential of this approach, compared to more conventional architectures designed to handle a single encoder, becomes apparent when the amount of data increases.

Effectiveness of molecular fingerprints for exploring the chemical space of natural products.

Natural products are a diverse class of compounds with promising biological properties, such as high potency and excellent selectivity. However, they have different structural motifs than typical drug-like compounds, e.g., a wider range of molecular weight, multiple stereocenters and higher fraction of sp3-hybridized carbons. This makes the encoding of natural products via molecular fingerprints difficult, thus restricting their use in cheminformatics studies. To tackle this issue, we explored over 30 years of research to systematically evaluate which molecular fingerprint provides the best performance on the natural product chemical space. We considered 20 molecular fingerprints from four different sources, which we then benchmarked on over 100,000 unique natural products from the COCONUT (COlleCtion of Open Natural prodUcTs) and CMNPD (Comprehensive Marine Natural Products Database) databases. Our analysis focused on the correlation between different fingerprints and their classification performance on 12 bioactivity prediction datasets. Our results show that different encodings can provide fundamentally different views of the natural product chemical space, leading to substantial differences in pairwise similarity and performance. While Extended Connectivity Fingerprints are the de-facto option to encoding drug-like compounds, other fingerprints resulted to match or outperform them for bioactivity prediction of natural products. These results highlight the need to evaluate multiple fingerprinting algorithms for optimal performance and suggest new areas of research. Finally, we provide an open-source Python package for computing all molecular fingerprints considered in the study, as well as data and scripts necessary to reproduce the results, at https://github.com/dahvida/NP_Fingerprints .

Quantitative structure-activity relationship models for ready biodegradability of chemicals.

The European REACH regulation requires information on ready biodegradation, which is a screening test to assess the biodegradability of chemicals. At the same time REACH encourages the use of alternatives to animal testing which includes predictions from quantitative structure-activity relationship (QSAR) models. The aim of this study was to build QSAR models to predict ready biodegradation of chemicals by using different modeling methods and types of molecular descriptors. Particular attention was given to data screening and validation procedures in order to build predictive models. Experimental values of 1055 chemicals were collected from the webpage of the National Institute of Technology and Evaluation of Japan (NITE): 837 and 218 molecules were used for calibration and testing purposes, respectively. In addition, models were further evaluated using an external validation set consisting of 670 molecules. Classification models were produced in order to discriminate biodegradable and nonbiodegradable chemicals by means of different mathematical methods: k nearest neighbors, partial least squares discriminant analysis, and support vector machines, as well as their consensus models. The proposed models and the derived consensus analysis demonstrated good classification performances with respect to already published QSAR models on biodegradation. Relationships between the molecular descriptors selected in each QSAR model and biodegradability were evaluated.

Classification-based machine learning approaches to predict the taste of molecules: A review.

The capacity to discriminate safe from dangerous compounds has played an important role in the evolution of species, including human beings. Highly evolved senses such as taste receptors allow humans to navigate and survive in the environment through information that arrives to the brain through electrical pulses. Specifically, taste receptors provide multiple bits of information about the substances that are introduced orally. These substances could be pleasant or not according to the taste responses that they trigger. Tastes have been classified into basic (sweet, bitter, umami, sour and salty) or non-basic (astringent, chilling, cooling, heating, pungent), while some compounds are considered as multitastes, taste modifiers or tasteless. Classification-based machine learning approaches are useful tools to develop predictive mathematical relationships in such a way as to predict the taste class of new molecules based on their chemical structure. This work reviews the history of multicriteria quantitative structure-taste relationship modelling, starting from the first ligand-based (LB) classifier proposed in 1980 by Lemont B. Kier and concluding with the most recent studies published in 2022.

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.

Response Surface Methodology Approach to Evaluate the Effect of Transition Metals and Oxygen on Photo-Degradation of Methionine in a Model Wine System Containing Riboflavin.

A Box-Behnken experimental design was implemented in model wine (MW) to clarify the impact of copper, iron, and oxygen in the photo-degradation of riboflavin (RF) and methionine (Met) by means of response surface methodology (RSM). Analogous experiments were undertaken in MW containing caffeic acid or catechin. The results evidenced the impact of copper, iron, and oxygen in the photo-induced reaction between RF and Met. In particular, considering a number of volatile sulfur compounds (VSCs) that act as markers of light-struck taste (LST), both transition metals can favor VSC formation, which was shown for the first time for iron. Oxygen in combination can also affect the concentration of VSCs, and a lower content of VSCs was revealed in the presence of phenols, especially caffeic acid. The perception of "cabbage" sensory character indicative of LST can be related to the transition metals as well as to the different phenols, with potentially strong prevention by phenolic acids.

NMR spectroscopy and chemometric models to detect a specific non-porcine ruminant contaminant in pharmaceutical heparin.

Heparin has been used successfully as a clinical antithrombotic for almost one century. Its isolation from animal sources (mostly porcine intestinal mucosa) involves multistep purification processes starting from the slaughterhouse (as mucosa) to the pharmaceutical plant (as the API). This complex supply chain increases the risk of contamination and adulteration, mainly with non-porcine ruminant material. The structural similarity of heparins from different origins, the natural variability of the heparin within samples from each source as well as the structural changes induced by manufacturing processes, require increasingly sophisticated methods capable of detecting low levels of contamination. The application of suitable multivariate classification approaches on API 1H NMRspectra serve as rapid and reliable tools for product authentication and the detection of contaminants. Soft Independent Modeling of Class Analogies (SIMCA), Discriminant Analysis (DA), Partial Least Square Discriminant Analysis (PLS-DA) and local classification methods (kNN, BNN and N3) were tested on about one hundred certified heparin samples produced by 14 different manufacturers revealing that Partial Least Squares Discriminant Analysis (PLS-DA) provided the best discrimination of contaminated batches, with a balanced accuracy of 97%.