Chiral pyrrolidines as multipotent agents in Alzheimer and neurodegenerative diseases.

In pharmaceutical science and drug design the versatility of the pyrrolidine scaffold relating to spatial arrangement, synthetic accessibility and pharmacological profile is a largely explored and most likely interesting one. Nonetheless, few evidences suggest the pivotal role of pyrrolidine as scaffold for multipotent agents in neurodegenerative diseases. We then challenged the enrolling in the field of Alzheimer disease of so far not ravelled targets of this chemical cliché with a structure based and computer-aided design strategy focusing on multi-target action, versatile synthesis as well as pharmacological safeness. To achieve these hits, ten enantiomeric pairs of compounds were obtained and tested, and the biological data will be here presented and discussed. Among the novel compounds, coumarin and sesamol scaffolds containing analogues resulted promising perspectives.

Nature-Inspired 1-Phenylpyrrolo[2,1-a]isoquinoline Scaffold for Novel Antiproliferative Agents Circumventing P-Glycoprotein-Dependent Multidrug Resistance.

Previous studies have shown that some lamellarin-resembling annelated azaheterocyclic carbaldehydes and related imino adducts, sharing the 1-phenyl-5,6-dihydropyrrolo[2,1-a]isoquinoline (1-Ph-DHPIQ) scaffold, are cytotoxic in some tumor cells and may reverse multidrug resistance (MDR) mediated by P-glycoprotein (P-gp). Herein, several novel substituted 1-Ph-DHPIQ derivatives were synthesized which carry carboxylate groups (COOH, COOEt), nitrile (CN) and Mannich bases (namely, morpholinomethyl derivatives) in the C2 position, as replacements of the already reported aldehyde group. They were evaluated for antiproliferative activity in four tumor cell lines (RD, HCT116, HeLa, A549) and for the ability of selectively inhibiting P-gp-mediated MDR. Lipophilicity descriptors and molecular docking calculations helped us in rationalizing the structure-activity relationships in the P-gp inhibition potency of the investigated 1-Ph-DHPIQs. As a main outcome, a morpholinomethyl Mannich base (8c) was disclosed which proved to be cytotoxic to all the tested tumor cell lines in the low micromolar range (IC50 < 20 µM) and to inhibit in vitro the efflux pumps P-gp and MRP1 responsible for MDR, with IC50s of 0.45 and 12.1 µM, respectively.

Novel 6-alkyl-bridged 4-arylalkylpiperazin-1-yl derivatives of azepino[4,3-b]indol-1(2H)-one as potent BChE-selective inhibitors showing protective effects against neurodegenerative insults.

Due to the putative role of butyrylcholinesterase (BChE) in regulation of acetylcholine levels and functions in the late stages of the Alzheimer's disease (AD), the potential of selective inhibitors (BChEIs) has been envisaged as an alternative to administration of acetylcholinesterase inhibitors (AChEIs). Starting from our recent findings, herein the synthesis and in vitro evaluation of cholinesterase (ChE) inhibition of a novel series of some twenty 3,4,5,6-tetrahydroazepino[4,3-b]indol-1(2H)-one derivatives, bearing at the indole nitrogen diverse alkyl-bridged 4-arylalkylpiperazin-1-yl chains, are reported. The length of the spacers, as well as the type of arylalkyl group affected the enzyme inhibition potency and BChE/AChE selectivity. Two compounds, namely 14c (IC50 = 163 nM) and 14d (IC50 = 65 nM), bearing at the nitrogen atom in position 6 a n-pentyl- or n-heptyl-bridged 4-phenethylpiperazin-1-yl chains, respectively, proved to be highly potent mixed-type inhibitors of both equine and human BChE isoforms, showing more than two order magnitude of selectivity over AChE. The study of binding kinetics through surface plasmon resonance (SPR) highlighted differences in their BChE residence times (8 and 47 s for 14c and 14d, respectively). Moreover, 14c and 14d proved to hit other mechanisms known to trigger neurodegeneration underlying AD and other CNS disorders. Unlike 14c, compound 14d proved also capable of inhibiting by more than 60% the in vitro self-induced aggregation of neurotoxic amyloid-ß (Aß) peptide at 100 µM concentration. On the other hand, 14c was slightly better than 14d in counteracting, at 1 and 10 µM concentration, glutamate excitotoxicity, due to over-excitation of NMDA receptors, and hydrogen peroxide-induced oxidative stress assessed in neuroblastoma cell line SH-SY5Y. This paper is dedicated to Prof. Marcello Ferappi, former dean of the Faculty of Pharmacy of the University of Bari, in the occasion of his 90th birthday.

Small Molecules for the Treatment of Long-COVID-Related Vascular Damage and Abnormal Blood Clotting: A Patent-Based Appraisal.

People affected by COVID-19 are exposed to, among others, abnormal clotting and endothelial dysfunction, which may result in deep vein thrombosis, cerebrovascular disorders, and ischemic and non-ischemic heart diseases, to mention a few. Treatments for COVID-19 include antiplatelet (e.g., aspirin, clopidogrel) and anticoagulant agents, but their impact on morbidity and mortality has not been proven. In addition, due to viremia-associated interconnected prothrombotic and proinflammatory events, anti-inflammatory drugs have also been investigated for their ability to mitigate against immune dysregulation due to the cytokine storm. By retrieving patent literature published in the last two years, small molecules patented for long-COVID-related blood clotting and hematological complications are herein examined, along with supporting evidence from preclinical and clinical studies. An overview of the main features and therapeutic potentials of small molecules is provided for the thromboxane receptor antagonist ramatroban, the pan-caspase inhibitor emricasan, and the sodium-hydrogen antiporter 1 (NHE-1) inhibitor rimeporide, as well as natural polyphenolic compounds.

Novel Dual-Acting Hybrids Targeting Type-2 Cannabinoid Receptors and Cholinesterase Activity Show Neuroprotective Effects In Vitro and Amelioration of Cognitive Impairment In Vivo.

Alzheimer's disease (AD) is a neurodegenerative form of dementia characterized by the loss of synapses and a progressive decline in cognitive abilities. Among current treatments for AD, acetylcholinesterase (AChE) inhibitors have efficacy limited to symptom relief, with significant side effects and poor compliance. Pharmacological agents that modulate the activity of type-2 cannabinoid receptors (CB2R) of the endocannabinoid system by activating or blocking them have also been shown to be effective against neuroinflammation. Herein, we describe the design, synthesis, and pharmacological effects in vitro and in vivo of dual-acting compounds that inhibit AChE and butyrylcholinesterase (BChE) and target CB2R. Within the investigated series, compound 4g proved to be the most promising. It achieved IC50 values in the low micromolar to submicromolar range against both human cholinesterase isoforms while antagonizing CB2R with Ki of 31 nM. Interestingly, 4g showed neuroprotective effects on the SH-SY5Y cell line thanks to its ability to prevent oxidative stress-induced cell toxicity and reverse scopolamine-induced amnesia in the Y-maze forced alternation test in vivo.

A second life for MAO inhibitors? From CNS diseases to anticancer therapy.

Monoamine oxidases A and B (MAO A, B) are ubiquitous enzymes responsible for oxidative deamination of amine neurotransmitters and xenobiotics. Despite decades of studies, MAO inhibitors (MAOIs) find today limited therapeutic space as second-line drugs for the treatment of depression and Parkinson's disease. In recent years, a renewed interest in MAOIs has been raised up by several studies investigating the role of MAOs, particularly MAO A, in tumor insurgence and progression, and the efficacy of MAOIs as coadjutants in the therapy of chemoresistant tumors. In this survey, we highlight the implication of MAOs in the biochemical pathways of tumorigenesis and review the state-of-the-art of preclinical and clinical studies of MAOIs as anticancer agents used in monotherapy or in combination with antitumor chemotherapeutics.

TISBE: A Public Web Platform for the Consensus-Based Explainable Prediction of Developmental Toxicity.

Despite being extremely relevant for the protection of prenatal and neonatal health, the developmental toxicity (Dev Tox) is a highly complex endpoint whose molecular rationale is still largely unknown. The lack of availability of high-quality data as well as robust nontesting methods makes its understanding even more difficult. Thus, the application of new explainable alternative methods is of utmost importance, with Dev Tox being one of the most animal-intensive research themes of regulatory toxicology. Descending from TIRESIA (Toxicology Intelligence and Regulatory Evaluations for Scientific and Industry Applications), the present work describes TISBE (TIRESIA Improved on Structure-Based Explainability), a new public web platform implementing four fundamental advancements for in silico analyses: a three times larger dataset, a transparent XAI (explainable artificial intelligence) framework employing a fragment-based fingerprint coding, a novel consensus classifier based on five independent machine learning models, and a new applicability domain (AD) method based on a double top-down approach for better estimating the prediction reliability. The training set (TS) includes as many as 1008 chemicals annotated with experimental toxicity values. Based on a 5-fold cross-validation, a median value of 0.410 for the Matthews correlation coefficient was calculated; TISBE was very effective, with a median value of sensitivity and specificity equal to 0.984 and 0.274, respectively. TISBE was applied on two external pools made of 1484 bioactive compounds and 85 pediatric drugs taken from ChEMBL (Chemical European Molecular Biology Laboratory) and TEDDY (Task-Force in Europe for Drug Development in the Young) repositories, respectively. Notably, TISBE gives users the option to clearly spot the molecular fragments responsible for the toxicity or the safety of a given chemical query and is available for free at https://prometheus.farmacia.uniba.it/tisbe.

A Critical Appraisal of the Protective Activity of Polyphenolic Antioxidants against Iatrogenic Effects of Anticancer Chemotherapeutics.

Polyphenolic compounds, encompassing flavonoids (e.g., quercetin, rutin, and cyanidin) and non-flavonoids (e.g., gallic acid, resveratrol, and curcumin), show several health-related beneficial effects, which include antioxidant, anti-inflammatory, hepatoprotective, antiviral, and anticarcinogenic properties, as well as the prevention of coronary heart diseases. Polyphenols have also been investigated for their counteraction against the adverse effects of common anticancer chemotherapeutics. This review evaluates the outcomes of clinical studies (and related preclinical data) over the last ten years, with a focus on the use of polyphenols in chemotherapy as auxiliary agents acting against oxidative stress toxicity induced by antitumor drugs. While further clinical studies are needed to establish adequate doses and optimal delivery systems, the improvement in polyphenols' metabolic stability and bioavailability, through the implementation of nanotechnologies that are currently being investigated, could improve therapeutic applications of their pharmaceutical or nutraceutical preparations in tumor chemotherapy.

Radiosynthesis and whole-body distribution in mice of a 18 F-labeled azepino[4,3-b]indole-1-one derivative with multimodal activity for the treatment of Alzheimer's disease.

Recently, the azepino[4,3-b]indole-1-one derivative 1 showed in vitro nanomolar inhibition against butyrylcholinesterase (BChE), the ChE isoform that plays a role in the progression and pathophysiology of Alzheimer's disease (AD), and protects against N-methyl- d-aspartate-induced neuronal toxicity. Three 9-R-substituted (R = F, Br, OMe) congeners were investigated. The 9-F derivative (2a) was found more potent as BChE inhibitors (half-maximal inhibitory concentration value = 21 nM) than 2b (9-Br) and 2c (9-OMe), achieving a residence time (38 s), assessed by surface plasmon resonance, threefold higher than that of 1. To progress in featuring the in vivo pharmacological characterization of 2a, herein the 18 F-labeled congener 2a was synthesized, by applying the aromatic 18 F-fluorination method, and its whole-body distribution in healthy mice, including brain penetration, was evaluated through positron emission tomography imaging. [18 F]2a exhibited a rapid and high brain uptake (3.35 ± 0.26% ID g-1 at 0.95 ± 0.15 min after injection), followed by a rapid clearance (t1/2 = 6.50 ± 0.93 min), showing good blood-brain barrier crossing. After a transient liver accumulation of [18 F]2a, the intestinal and urinary excretion was quantified. Finally, ex vivo pharmacological experiments in mice showed that the unlabeled 2a affects the transmitters' neurochemistry, which might be favorable to reverse cognition impairment in mild-to-moderate AD-related dementias.

Making sense of chemical space network shows signs of criticality.

Chemical space modelling has great importance in unveiling and visualising latent information, which is critical in predictive toxicology related to drug discovery process. While the use of traditional molecular descriptors and fingerprints may suffer from the so-called curse of dimensionality, complex networks are devoid of the typical drawbacks of coordinate-based representations. Herein, we use chemical space networks (CSNs) to analyse the case of the developmental toxicity (Dev Tox), which remains a challenging endpoint for the difficulty of gathering enough reliable data despite very important for the protection of the maternal and child health. Our study proved that the Dev Tox CSN has a complex non-random organisation and can thus provide a wealth of meaningful information also for predictive purposes. At a phase transition, chemical similarities highlight well-established toxicophores, such as aryl derivatives, mostly neurotoxic hydantoins, barbiturates and amino alcohols, steroids, and volatile organic compounds ether-like chemicals, which are strongly suspected of the Dev Tox onset and can thus be employed as effective alerts for prioritising chemicals before testing.

Where developmental toxicity meets explainable artificial intelligence: state-of-the-art and perspectives.

INTRODUCTION: The application of Artificial Intelligence (AI) to predictive toxicology is rapidly increasing, particularly aiming to develop non-testing methods that effectively address ethical concerns and reduce economic costs. In this context, Developmental Toxicity (Dev Tox) stands as a key human health endpoint, especially significant for safeguarding maternal and child well-being. AREAS COVERED: This review outlines the existing methods employed in Dev Tox predictions and underscores the benefits of utilizing New Approach Methodologies (NAMs), specifically focusing on eXplainable Artificial Intelligence (XAI), which proves highly efficient in constructing reliable and transparent models aligned with recommendations from international regulatory bodies. EXPERT OPINION: The limited availability of high-quality data and the absence of dependable Dev Tox methodologies render XAI an appealing avenue for systematically developing interpretable and transparent models, which hold immense potential for both scientific evaluations and regulatory decision-making.

In-vitro and in-silico studies of annelated 1,4,7,8-tetrahydroazocine ester derivatives as nanomolar selective inhibitors of human butyrylcholinesterase.

Based on previous finding showing 2,3,6,11-tetrahydro-1H-azocino[4,5-b]indole as suitable scaffold of novel inhibitors of acetylcholinesterase (AChE), a main target of drugs for the treatment of Alzheimer's disease and related dementias, herein we investigated diverse newly and previously synthesized ß-enamino esters (and ketones) derivatives of 1,4,7,8-tetrahydroazocines (and some azonines) fused with benzene, 1H-indole, 4H-chromen-4-one and pyrimidin-4(3H)-one. Twenty derivatives of diversely annelated eight-to-nine-membered azaheterocyclic ring, prepared through domino reaction of the respective tetrahydropyridine and azepine with activated alkynes, were assayed for the inhibitory activity against AChE and butyrylcholinesterase (BChE). As a major outcome, compound 7c, an alkylamino derivative of tetrahydropyrimido[4,5-d]azocine, was found to be a highly potent BChE-selective inhibitor, which showed a noncompetitive/mixed-type inhibition mechanism against human BChE with single digit nanomolar inhibition constant (Ki = 7.8 ± 0.2 nM). The four-order magnitude BChE-selectivity of 7c clearly reflects the effect of lipophilicity upon binding to the BChE binding cavity. The ChEs' inhibition data, interpreted by chemoinformatic tools and an in-depth in-silico study (molecular docking combined with molecular dynamics calculations), not only highlighted key structural factors enhancing inhibition potency and selectivity toward BChE, but also shed light on subtle differences distinguishing the binding sites of equine BChE from the recombinant human BChE. Compound 7c inhibited P-glycoprotein with IC50 of 0.27 µM, which may support its ability to permeate blood-brain barrier, and proved to be no cytotoxic in human liver cancer cell line (HepG2) at the BChE bioactive concentrations. Overall, the biological profile allows us to envision 7c as a promising template to improve design and development of BChE-selective ligands of pharmaceutical interest, including inhibitors and fluorogenic probes.

CIRCE: Web-Based Platform for the Prediction of Cannabinoid Receptor Ligands Using Explainable Machine Learning.

The endocannabinoid system, which includes cannabinoid receptor 1 and 2 subtypes (CB1R and CB2R, respectively), is responsible for the onset of various pathologies including neurodegeneration, cancer, neuropathic and inflammatory pain, obesity, and inflammatory bowel disease. Given the high similarity of CB1R and CB2R, generating subtype-selective ligands is still an open challenge. In this work, the Cannabinoid Iterative Revaluation for Classification and Explanation (CIRCE) compound prediction platform has been generated based on explainable machine learning to support the design of selective CB1R and CB2R ligands. Multilayer classifiers were combined with Shapley value analysis to facilitate explainable predictions. In test calculations, CIRCE predictions reached ∼80% accuracy and structural features determining ligand predictions were rationalized. CIRCE was designed as a web-based prediction platform that is made freely available as a part of our study.

Investigation on Novel E/Z 2-Benzylideneindan-1-One-Based Photoswitches with AChE and MAO-B Dual Inhibitory Activity.

The multitarget therapeutic strategy, as opposed to the more traditional 'one disease-one target-one drug', may hold promise in treating multifactorial neurodegenerative syndromes, such as Alzheimer's disease (AD) and related dementias. Recently, combining a photopharmacology approach with the multitarget-directed ligand (MTDL) design strategy, we disclosed a novel donepezil-like compound, namely 2-(4-((diethylamino)methyl)benzylidene)-5-methoxy-2,3-dihydro-1H-inden-1-one (1a), which in the E isomeric form (and about tenfold less in the UV-B photo-induced isomer Z) showed the best activity as dual inhibitor of the AD-related targets acetylcholinesterase (AChE) and monoamine oxidase B (MAO-B). Herein, we investigated further photoisomerizable 2-benzylideneindan-1-one analogs 1b-h with the unconjugated tertiary amino moiety bearing alkyls of different bulkiness and lipophilicity. For each compound, the thermal stable E geometric isomer, along with the E/Z mixture as produced by UV-B light irradiation in the photostationary state (PSS, 75% Z), was investigated for the inhibition of human ChEs and MAOs. The pure E-isomer of the N-benzyl(ethyl)amino analog 1h achieved low nanomolar AChE and high nanomolar MAO-B inhibition potencies (IC50s 39 and 355 nM, respectively), whereas photoisomerization to the Z isomer (75% Z in the PSS mixture) resulted in a decrease (about 30%) of AChE inhibitory potency, and not in the MAO-B one. Molecular docking studies were performed to rationalize the different E/Z selectivity of 1h toward the two target enzymes.

Synthesis, computational and experimental pharmacological studies for (thio)ether-triazine 5-HT6R ligands with noticeable action on AChE/BChE and chalcogen-dependent intrinsic activity in search for new class of drugs against Alzheimer's disease.

Alzheimer's disease is becoming a growing problem increasing at a tremendous rate. Serotonin 5-HT6 receptors appear to be a particularly attractive target from a therapeutic perspective, due to their involvement not only in cognitive processes, but also in depression and psychosis. In this work, we present the synthesis and broad biological characterization of a new series of 18 compounds with a unique 1,3,5-triazine backbone, as potent 5-HT6 receptor ligands. The main aim of this research is to compare the biological activity of the newly synthesized sulfur derivatives with their oxygen analogues and their N-demethylated O- and S-metabolites obtained for the first time. Most of the new triazines displayed high affinity (Ki < 200 nM) and selectivity towards 5-HT6R, with respect to 5-HT2AR, 5-HT7R, and D2R, in the radioligand binding assays. For selected, active compounds crystallographic studies, functional bioassays, and ADME-Tox profile in vitro were performed. The exciting novelty is that the sulfur derivatives exhibit an agonistic mode of action contrary to all other compounds obtained to date in this chemical class herein and previously reported. Advanced computational studies indicated that this intriguing functional shift might be caused by presence of chalcogen bonds formed only by the sulfur atom. In addition, the N-demethylated derivatives have emerged highly potent antioxidants and, moreover, show a significant improvement in metabolic stability compared to the parent structures. The cholinesterase study present micromolar inhibitory AChE and BChE activity for both 5-HT6 agonist 19 and potent antagonist 5. Finally, the behavioral experiments of compound 19 demonstrated its antidepressant-like properties and slight ability to improve cognitive deficits, without inducing memory impairments by itself. Described pharmacological properties of both compounds (5 and 19) allow to give a design clue for the development of multitarget compounds with 5-HT6 (both agonist and antagonist)/AChE and/or BChE mechanism in the group of 1,3,5-triazine derivatives.

Short-term anti-remodeling effects of gliflozins in diabetic patients with heart failure and reduced ejection fraction: an explainable artificial intelligence approach.

Introduction: Sodium-glucose cotransporter type 2 inhibitors (SGLT2i), gliflozins, play an emerging role for the treatment of heart failure with reduced left ventricular ejection fraction (HFrEF). Nevertheless, the effects of SGLT2i on ventricular remodeling and function have not been completely understood yet. Explainable artificial intelligence represents an unprecedented explorative option to clinical research in this field. Based on echocardiographic evaluations, we identified some key clinical responses to gliflozins by employing a machine learning approach. Methods: Seventy-eight consecutive diabetic outpatients followed for HFrEF were enrolled in the study. Using a random forests classification, a single subject analysis was performed to define the profile of patients treated with gliflozins. An explainability analysis using Shapley values was used to outline clinical parameters that mostly improved after gliflozin therapy and machine learning runs highlighted specific variables predictive of gliflozin response. Results: The five-fold cross-validation analyses showed that gliflozins patients can be identified with a 0.70 ± 0.03% accuracy. The most relevant parameters distinguishing gliflozins patients were Right Ventricular S'-Velocity, Left Ventricular End Systolic Diameter and E/e' ratio. In addition, low Tricuspid Annular Plane Systolic Excursion values along with high Left Ventricular End Systolic Diameter and End Diastolic Volume values were associated to lower gliflozin efficacy in terms of anti-remodeling effects. Discussion: In conclusion, a machine learning analysis on a population of diabetic patients with HFrEF showed that SGLT2i treatment improved left ventricular remodeling, left ventricular diastolic and biventricular systolic function. This cardiovascular response may be predicted by routine echocardiographic parameters, with an explainable artificial intelligence approach, suggesting a lower efficacy in case of advanced stages of cardiac remodeling.

Chemical and Biological Evaluation of Novel 1H-Chromeno[3,2-c]pyridine Derivatives as MAO Inhibitors Endowed with Potential Anticancer Activity.

About twenty molecules sharing 1H-chromeno[3,2-c]pyridine as the scaffold and differing in the degree of saturation of the pyridine ring, oxidation at C10, 1-phenylethynyl at C1 and 1H-indol-3-yl fragments at C10, as well as a few small substituents at C6 and C8, were synthesized starting from 1,2,3,4-tetrahydro-2-methylchromeno[3,2-c]pyridin-10-ones (1,2,3,4-THCP-10-ones, 1) or 2,3-dihydro-2-methyl-1H-chromeno[3,2-c]pyridines (2,3-DHPCs, 2). The newly synthesized compounds were tested as inhibitors of the human isoforms of monoamine oxidase (MAO A and B) and cholinesterase (AChE and BChE), and the following main SARs were inferred: (i) The 2,3-DHCP derivatives 2 inhibit MAO A (IC50 about 1 µM) preferentially; (ii) the 1,2,3,4-THCP-10-one 3a, bearing the phenylethynyl fragment at C1, returned as a potent MAO B inhibitor (IC50 0.51 µM) and moderate inhibitor of both ChEs (IC50s 7-8 µM); (iii) the 1H-indol-3-yl fragment at C10 slightly increases the MAO B inhibition potency, with the analog 6c achieving MAO B IC50 of 3.51 µM. The MAO B inhibitor 3a deserves further pharmacological studies as a remedy in the symptomatic treatment of Parkinson's disease and neuroprotectant for Alzheimer's disease. Besides the established neuroprotective effects of MAO inhibitors, the role of MAOs in tumor insurgence and progression has been recently reported. Herein, antiproliferative assays with breast (MCF-7), colon (HCT116) and cisplatin-resistant ovarian (SK-OV-3) tumor cells revealed that the 10-indolyl-bearing 2,3,4,10-THCP analog 6c exerts anti-tumor activity with IC50s in the range 4.83-11.3 µM.

Synthesis of Novel Benzo[b][1,6]naphthyridine Derivatives and Investigation of Their Potential as Scaffolds of MAO Inhibitors.

In this work, 2-alkyl-10-chloro-1,2,3,4-tetrahydrobenzo[b][1,6]naphthyridines were obtained and their reactivity was studied. Novel derivatives of the tricyclic scaffold, including 1-phenylethynyl (5), 1-indol-3-yl (8), and azocino[4,5-b]quinoline (10) derivatives, were synthesized and characterized herein for the first time. Among the newly synthesized derivatives, 5c-h proved to be MAO B inhibitors with potency in the low micromolar range. In particular, the 1-(2-(4-fluorophenyl)ethynyl) analog 5g achieved an IC50 of 1.35 µM, a value close to that of the well-known MAO B inhibitor pargyline.

TIRESIA: An eXplainable Artificial Intelligence Platform for Predicting Developmental Toxicity.

Herein, a robust and reproducible eXplainable Artificial Intelligence (XAI) approach is presented, which allows prediction of developmental toxicity, a challenging human-health endpoint in toxicology. The application of XAI as an alternative method is of the utmost importance with developmental toxicity being one of the most animal-intensive areas of regulatory toxicology. In this work, the established CAESAR (Computer Assisted Evaluation of industrial chemical Substances According to Regulations) training set made of 234 chemicals for model learning is employed. Two test sets, including as a whole 585 chemicals, were instead used for validation and generalization purposes. The proposed framework favorably compares with the state-of-the-art approaches in terms of accuracy, sensitivity, and specificity, thus resulting in a reliable support system for developmental toxicity ensuring informativeness, uncertainty estimation, generalization, and transparency. Based on the eXtreme Gradient Boosting (XGB) algorithm, our predictive model provides easy interpretative keys based on specific molecular descriptors and structural alerts enabling one to distinguish toxic and nontoxic chemicals. Inspired by the Organisation for Economic Co-operation and Development (OECD) principles for the validation of Quantitative Structure-Activity Relationships (QSARs) for regulatory purposes, the results are summarized in a standard report in portable document format, enclosing also details concerned with a density-based model applicability domain and SHAP (SHapley Additive exPlanations) explainability, the latter particularly useful to better understand the effective roles played by molecular features. Notably, our model has been implemented in TIRESIA (Toxicology Intelligence and Regulatory Evaluations for Scientific and Industry Applications), a free of charge web platform available at http://tiresia.uniba.it.