Design of New Dispersants Using Machine Learning and Visual Analytics.

Artificial intelligence (AI) is an emerging technology that is revolutionizing the discovery of new materials. One key application of AI is virtual screening of chemical libraries, which enables the accelerated discovery of materials with desired properties. In this study, we developed computational models to predict the dispersancy efficiency of oil and lubricant additives, a critical property in their design that can be estimated through a quantity named blotter spot. We propose a comprehensive approach that combines machine learning techniques with visual analytics strategies in an interactive tool that supports domain experts' decision-making. We evaluated the proposed models quantitatively and illustrated their benefits through a case study. Specifically, we analyzed a series of virtual polyisobutylene succinimide (PIBSI) molecules derived from a known reference substrate. Our best-performing probabilistic model was Bayesian Additive Regression Trees (BART), which achieved a mean absolute error of 5.50±0.34 and a root mean square error of 7.56±0.47, as estimated through 5-fold cross-validation. To facilitate future research, we have made the dataset, including the potential dispersants used for modeling, publicly available. Our approach can help accelerate the discovery of new oil and lubricant additives, and our interactive tool can aid domain experts in making informed decisions based on blotter spot and other key properties.

Multitask Deep Neural Networks for Ames Mutagenicity Prediction.

The Ames mutagenicity test constitutes the most frequently used assay to estimate the mutagenic potential of drug candidates. While this test employs experimental results using various strains of Salmonella typhimurium, the vast majority of the published in silico models for predicting mutagenicity do not take into account the test results of the individual experiments conducted for each strain. Instead, such QSAR models are generally trained employing overall labels (i.e., mutagenic and nonmutagenic). Recently, neural-based models combined with multitask learning strategies have yielded interesting results in different domains, given their capabilities to model multitarget functions. In this scenario, we propose a novel neural-based QSAR model to predict mutagenicity that leverages experimental results from different strains involved in the Ames test by means of a multitask learning approach. To the best of our knowledge, the modeling strategy hereby proposed has not been applied to model Ames mutagenicity previously. The results yielded by our model surpass those obtained by single-task modeling strategies, such as models that predict the overall Ames label or ensemble models built from individual strains. For reproducibility and accessibility purposes, all source code and datasets used in our experiments are publicly available.

Computational Drug Repositioning for Chagas Disease Using Protein-Ligand Interaction Profiling.

Chagas disease, caused by Trypanosoma cruzi (T. cruzi), affects nearly eight million people worldwide. There are currently only limited treatment options, which cause several side effects and have drug resistance. Thus, there is a great need for a novel, improved Chagas treatment. Bifunctional enzyme dihydrofolate reductase-thymidylate synthase (DHFR-TS) has emerged as a promising pharmacological target. Moreover, some human dihydrofolate reductase (HsDHFR) inhibitors such as trimetrexate also inhibit T. cruzi DHFR-TS (TcDHFR-TS). These compounds serve as a starting point and a reference in a screening campaign to search for new TcDHFR-TS inhibitors. In this paper, a novel virtual screening approach was developed that combines classical docking with protein-ligand interaction profiling to identify drug repositioning opportunities against T. cruzi infection. In this approach, some food and drug administration (FDA)-approved drugs that were predicted to bind with high affinity to TcDHFR-TS and whose predicted molecular interactions are conserved among known inhibitors were selected. Overall, ten putative TcDHFR-TS inhibitors were identified. These exhibited a similar interaction profile and a higher computed binding affinity, compared to trimetrexate. Nilotinib, glipizide, glyburide and gliquidone were tested on T. cruzi epimastigotes and showed growth inhibitory activity in the micromolar range. Therefore, these compounds could lead to the development of new treatment options for Chagas disease.

QSAR Classification Models for Predicting the Activity of Inhibitors of Beta-Secretase (BACE1) Associated with Alzheimer's Disease.

Alzheimer's disease is one of the most common neurodegenerative disorders in elder population. The ß-site amyloid cleavage enzyme 1 (BACE1) is the major constituent of amyloid plaques and plays a central role in this brain pathogenesis, thus it constitutes an auspicious pharmacological target for its treatment. In this paper, a QSAR model for identification of potential inhibitors of BACE1 protein is designed by using classification methods. For building this model, a database with 215 molecules collected from different sources has been assembled. This dataset contains diverse compounds with different scaffolds and physical-chemical properties, covering a wide chemical space in the drug-like range. The most distinctive aspect of the applied QSAR strategy is the combination of hybridization with backward elimination of models, which contributes to improve the quality of the final QSAR model. Another relevant step is the visual analysis of the molecular descriptors that allows guaranteeing the absence of information redundancy in the model. The QSAR model performances have been assessed by traditional metrics, and the final proposed model has low cardinality, and reaches a high percentage of chemical compounds correctly classified.

Highly potent and selective aryl-1,2,3-triazolyl benzylpiperidine inhibitors toward butyrylcholinesterase in Alzheimer's disease.

Acetylcholinesterase (AChE) is the key enzyme targeted in Alzheimer's disease (AD) therapy, nevertheless butyrylcholinesterase (BuChE) has been drawing attention due to its role in the disease progression. Thus, we aimed to synthesize novel cholinesterases inhibitors considering structural differences in their peripheral site, exploiting a moiety replacement approach based on the potent and selective hAChE drug donepezil. Hence, two small series of N-benzylpiperidine based compounds have successfully been synthesized as novel potent and selective hBuChE inhibitors. The most promising compounds (9 and 11) were not cytotoxic and their kinetic study accounted for dual binding site mode of interaction, which is in agreement with further docking and molecular dynamics studies. Therefore, this study demonstrates how our strategy enabled the discovery of novel promising and privileged structures. Remarkably, compound 11 proved to be one of the most potent (0.17 nM) and selective (>58,000-fold) hBuChE inhibitor ever reported.

Chameleon-like behavior of indolylpiperidines in complex with cholinesterases targets: Potent butyrylcholinesterase inhibitors.

Alzheimer's disease (AD) is the most common form of dementia worldwide with an increasing prevalence for the next years. The multifactorial nature of AD precludes the design of new drugs directed to a single target being probably one of the reasons for recent failures. Therefore, dual binding site acetylcholinesterase (AChE) inhibitors have been revealed as cognitive enhancers and ß-amyloid modulators offering an alternative in AD therapy field. Based on the dual ligands NP61 and donepezil, the present study reports the synthesis of a series of indolylpiperidines hybrids to optimize the NP61 structure preserving the indole nucleus, but replacing the tacrine moiety of NP61 by benzyl piperidine core found in donepezil. Surprisingly, this new family of indolylpiperidines derivatives showed very potent and selective hBuChE inhibition. Further studies of NMR and molecular dynamics have showed the capacity of these hybrid molecules to change their bioactive conformation depending on the binding site, being capable to inhibit with different shapes BuChE and residually AChE.

From dual binding site acetylcholinesterase inhibitors to allosteric modulators: A new avenue for disease-modifying drugs in Alzheimer's disease.

The lack of an effective treatment for Alzheimer' disease (AD), an increasing prevalence and severe neurodegenerative pathology boost medicinal chemists to look for new drugs. Currently, only acethylcholinesterase (AChE) inhibitors and glutamate antagonist have been approved to the palliative treatment of AD. Although they have a short-term symptomatic benefits, their clinical use have revealed important non-cholinergic functions for AChE such its chaperone role in beta-amyloid toxicity. We propose here the design, synthesis and evaluation of non-toxic dual binding site AChEIs by hybridization of indanone and quinoline heterocyclic scaffolds. Unexpectely, we have found a potent allosteric modulator of AChE able to target cholinergic and non-cholinergic functions by fixing a specific AChE conformation, confirmed by STD-NMR and molecular modeling studies. Furthermore the promising biological data obtained on human neuroblastoma SH-SY5Y cell assays for the new allosteric hybrid 14, led us to propose it as a valuable pharmacological tool for the study of non-cholinergic functions of AChE, and as a new important lead for novel disease modifying agents against AD.

Hybridizing Feature Selection and Feature Learning Approaches in QSAR Modeling for Drug Discovery.

Quantitative structure-activity relationship modeling using machine learning techniques constitutes a complex computational problem, where the identification of the most informative molecular descriptors for predicting a specific target property plays a critical role. Two main general approaches can be used for this modeling procedure: feature selection and feature learning. In this paper, a performance comparative study of two state-of-art methods related to these two approaches is carried out. In particular, regression and classification models for three different issues are inferred using both methods under different experimental scenarios: two drug-like properties, such as blood-brain-barrier and human intestinal absorption, and enantiomeric excess, as a measurement of purity used for chiral substances. Beyond the contrastive analysis of feature selection and feature learning methods as competitive approaches, the hybridization of these strategies is also evaluated based on previous results obtained in material sciences. From the experimental results, it can be concluded that there is not a clear winner between both approaches because the performance depends on the characteristics of the compound databases used for modeling. Nevertheless, in several cases, it was observed that the accuracy of the models can be improved by combining both approaches when the molecular descriptor sets provided by feature selection and feature learning contain complementary information.

In vivo anti-Chagas vinylthio-, vinylsulfinyl-, and vinylsulfonylbenzofuroxan derivatives.

New benzofuroxans were developed and studied as antiproliferative Trypanosoma cruzi agents. Compounds displayed remarkable in vitro activities against different strains, Tulahuen 2, CL Brener and Y. Its unspecific cytotoxicity was evaluated using human macrophages being not toxic at a concentration at least 8 times, and until 250 times, that of its T. cruzi IC50. Some biochemical pathways were studied, namely parasite respiration, cysteinyl active site enzymes and reaction with glutathione, as target for the mechanism of action. Not only T. cruzi respiration but also Cruzipain or trypanothione reductase were not affected, however the most active derivatives, the vinylsulfinyl- and vinylsulfonyl-containing benzofuroxans, react with glutathione in a redox pathway. Furthermore, the compounds showed good in vivo activities when they were studied in an acute murine model of Chagas' disease. The compounds were able to reduce the parasite loads of animals with fully established T. cruzi infections.

Synthesis and nematocide activity of S-glycopyranosyl-6,7-diarylthiolumazines.

6,7-Diaryl derivatives of mono and di-S-glycopyranosylthiolumazine derivatives 5-8 were prepared to test their nematocide activity. In vitro tests against Caenorhabditis elegans were performed and it was found that monosubstituted derivatives 5-7 showed higher activity than the corresponding unsubstituted 2-thiolumazines 1-3, whilst 2-S,4-S-di-glycopyranosylpteridine derivative 8 was inactive in contrast to unsubstituted derivative 4. In order to check whether the lack of activity of 8 was due to the two bulky substituents of the pteridine nucleus, 2-S,4-S-dimethyl derivative 9 was synthesized and assayed showing also lack of activity. A theoretical study on the stability of the different possible tautomers of compound 4 was carried out in an attempt to explain some, in appearance, anomalous (13)C NMR data of this compound.