In this study, we repositioned thiopurine drugs used for the treatment of acute leukaemia as new tyrosinase inhibitors. Tyrosinase catalyses two distinct and successive oxidations in melanin biosynthesis: the conversions of tyrosine to dihydroxyphenylalanine (DOPA) and DOPA to dopaquinone. Continuous efforts are underway to discover small molecule inhibitors of tyrosinase for therapeutic, cosmetic, and agricultural purposes. Structure-based virtual screening has predicted inhibitor candidates for mushroom tyrosinase from drugs approved by the US Food and Drug Administration (FDA). Enzyme assays have confirmed the thiopurine leukaemia drug, thioguanine, as a tyrosinase inhibitor. Two other thiopurine drugs, mercaptopurine and azathioprine, were also evaluated for their tyrosinase inhibitory activity; mercaptopurine caused stronger inhibition than thioguanine did, whereas azathioprine was a poor inhibitor. The inhibitory constants of thioguanine and mercaptopurine were calculated as 52 and 16 µM, respectively, and the value of mercaptopurine was comparable to that of the well-known inhibitor kojic acid (13 µM). The cell lysate and melanin content assay in B16F10 melanoma cells confirmed that the compounds inhibited mammalian tyrosinase. In particular, 50 µM thioguanine reduced the melanin content by 57% without cytotoxicity. Furthermore, the thiopurine drugs shared little chemical similarity with the known tyrosinase inhibitors.

We discuss further details on the concepts of “drug-likeness”, “lead-likeness”, and “natural product-likeness”. The discussion will first focus on natural products as drugs, then a discussion of previous studies in which the complexities of the scaffolds and chemical space of naturally occurring compounds have been compared with synthetic, semi-synthetic compounds and FDA-approved drugs. This is followed by guiding principles for designing “drug-like” natural product libraries for lead compound discovery purposes. We end up by presenting a tool for measuring “natural product-likeness” of compounds and a brief presentation of machine learning approaches and a binary quantitative structure-activity relationship (QSAR) for classifying drugs from non-drugs and natural compounds from non-natural ones, respectively.

Non-systemic oral drugs, which do not get absorbed through the gastrointestinal membrane, are an important group of compounds for targeting diseases residing within the gastrointestinal lumen. However, they require a set of design principles not sufficiently covered by traditional druglikeness metrics, such as Lipinski’s Rule of 5. These druglikeness metrics fail to accurately identify true negative outcomes; such is the requirement for novel cheminformatic based approaches for the discovery of non-systemic small molecules. One such example of these newer metrics is the fraction of SP3-hybridised carbon atoms (fSP3) which has shown promising application as an identifier of poor oral-bioavailability. Herein, we discuss the application of fSP3 in the drug discovery process in a number of case studies, including our own work discovering non-systemic fructose scavengers for the treatment of fructose malabsorption.

Taste determination in small molecules is critical in food chemistry, but traditional experimental methods can be time-consuming. Consequently, computational techniques have emerged as val-uable tools for this task. In this study, we explore taste prediction using various molecular feature representations and assess the performance of different machine learning algorithms on a dataset comprising 2,601 molecules. The results reveal that GNN-based models outperform other ap-proaches in taste prediction. Moreover, consensus models that combine diverse molecular repre-sentations demonstrate improved performance. Among these, molecular fingerprints + GNN con-sensus model emerges as the top performer, highlighting the complementary strengths of GNNs and molecular fingerprints. These findings have significant implications for food chemistry research and related fields. By leveraging these computational approaches, taste prediction can be expedited, leading to advancements in understanding the relationship between molecular structure and taste perception in various food components and related compounds.

Molecules sourced from marine environments hold immense promise for the development of novel therapeutic drugs, owing to their distinctive chemical compositions and valuable medicinal attributes. Notably, Talarolide A and Talaropeptides A-D have gained recent attention as potential candidates for pharmaceutical applications. This study aims to explore the chemical reactivity of Talarolide A and Talaropeptides A-D through the application of molecular modeling and computational chemistry techniques, specifically employing Conceptual Density Functional Theory (CDFT). By investigating their chemical behaviors, the study seeks to contribute to the understanding of the potential pharmacological uses of these marine-derived compounds. The investigation involves the utilization of molecular modeling and computational chemistry tools, with a specific focus on Conceptual DFT. These techniques are applied to examine the chemical reactivity of Talarolide A and Talaropeptides A-D, shedding light on their potential for interacting with various biological targets. The research reveals insights into the chemical reactivity of Talarolide A and Talaropeptides A-D, uncovering active reaction sites and potential biological targets. Talarolide A demonstrates robust anti-inflammatory properties and antitumor activity against diverse cancer cell lines, while Talaropeptide A displays antimicrobial efficacy against both bacteria and fungi. The study underscores the significance of comprehending the chemical reactivity of natural marine products for advancing pharmaceutical development. By employing computational tools such as Conceptual DFT, this research uncovers crucial information about the active sites and potential targets of Talarolide A and Talaropeptides A-D. The findings contribute to the broader understanding of marine-derived peptides and their prospects as pharmaceutical agents, though further investigations are necessary to fully elucidate their pharmacological attributes.

The question of molecular similarity is core in cheminformatics, and is usually assessed via a pairwise comparison based on vectors of properties or molecular fingerprints. We recently exploited variational autoencoders to embed 6M molecules in a chemical space, such that their (Euclidean) distance within the latent space so formed could be assessed within the framework of the entire molecular set. However, the standard objective function used did not seek to manipulate the latent space so as to cluster the molecules based on any perceived similarity. Using a set of some 160,000 molecules of biological relevance, we here bring together three modern elements of deep learning to create a novel and disentangled latent space, viz transformers, contrastive learning, and an embedded autoencoder. The effective dimensionality of the latent space was varied such that clear separation of individual types of molecules could be observed within individual dimensions of the latent space. The capacity of the network was such that many dimensions were not populated at all. As before, we assessed the utility of the representation by comparing clozapine with its near neighbours, and did the same for various antibiotics related to flucloxacillin. Transformers, especially when as here coupled with contrastive learning, effectively provide one-shot learning, and lead to a successful and disentangled representation of molecular latent spaces that at once uses the entire training set in their construction while allowing ‘similar’ molecules to cluster together in an effective and interpretable way.

We report the isolation of two novel degrading bacterial strains active against the potential environmental pollutant acetamino-phen/paracetamol. Streptomyces Chrestomyceticus (RS2) and Streptomyces Flavofuscus (M33) collected from El-Natrun Valley, Egypt, water, sediment, and sand samples were taxonomically characterized via conidiophores morphologic characterization under Trans-mission Electron Microscope (TEM). Genotypic identification, done based on their 16S rRNA gene sequence analysis followed by BLAST alignment and deposited on the NCBI as 2 novel strains. The phylogenetic tree was constructed using an appropriate software. Acetaminophen degradation-products chemical structure was identified by GC/LC MS. The biological antimicrobial activity of some selected acetaminophen degradation-products extracts and derived compounds was examined against panel of test micro-organisms and found to be have high anti-microbial effect. In addition, in silico cheminformatics Swiss ADMET evaluation of these selected deg-radation extract for gastric absorption, distribution, hepatic metabolism, renal excretion as well as distribution to CNS and finally not mutagenic/teratogenic or genotoxic virtually. In vitro cytotoxic activity of these selected bio-degradation products was performed against HepG2 and MCF7 cancer cell lines, revealing that M33 and RS2 extract effect on acetaminophen/paracetamol bio-degradation products are potentially safe with higher IC50 on HepG2 and MCF7, in comparison to acetaminophen/paracetamol with IC50 of 108.5 μg/ml. Moreover, the in vivo acute oral single dose toxicity experiment was conducted, being compared with acetamino-phen/paracetamol to confirm that the bio-degradation products have lower toxicity than acetaminophen in vivo and in silico.

Lead optimization in drug discovery is a crucial phase where initial hits are refined into compounds with improved pharmacological properties. While traditional methods rely on manual experimentation and modifications, AI-driven techniques have revolutionized this process by leveraging big data and predictive modeling. This review explores how AI-driven approaches accelerate lead optimization, showcasing examples like deep neural networks and reinforcement learning. Integration of multi-omic data and experimental validation further enhances AI-driven strategies. The future lies in refining these AI methods, democratizing tools, and interdisciplinary collaboration to streamline drug discovery and address medical needs efficiently.

In the realm of modern drug discovery, the integration of chemoinformatics and quantitative structure-activity relationship (QSAR) modeling has emerged as a formidable alliance, enabling researchers to harness the vast potential of machine learning (ML) techniques for predictive molecular design and analysis. This review delves into the fundamental aspects of chemoinformatics, elucidating the intricate nature of chemical data and the crucial role of molecular descriptors in unveiling the underlying molecular properties. Molecular descriptors, including 2D fingerprints and topological indices, in conjunction with the structural-activity relationships (SARs), play a pivotal role in unlocking the pathway to small-molecule drug discovery. Technical intricacies of developing robust ML-QSAR models, including feature selection, model validation, and performance evaluation, are discussed herewith. Various ML algorithms, such as regression analysis and support vector machines, are showcased in the text for their ability to predict and comprehend the relationships between molecular structures and biological activities. This review serves as a comprehensive guide for researchers, providing an understanding of the synergy between chemoinformatics, QSAR, and ML. By embracing these cutting-edge technologies, predictive molecular analysis holds promise for expediting the discovery of novel therapeutic agents in the pharmaceutical sciences.

The accurate prediction of molecular properties such as lipophilicity and aqueous solubility are of great importance and pose challenges in several stages of the drug discovery pipeline. Machine learning methods like graph-based neural networks (GNNs) have shown exceptionally good performance in predicting these properties. In this work, we introduce a novel GNN architecture, called directed edge graph isomorphism network (D-GIN). It is composed of two distinct sub-architectures (D-MPNN, GIN) and achieves an improvement in accuracy over its sub-architectures employing various learning, and featurization strategies. We argue that combining models with different key aspects help make graph neural networks deeper and simultaneously increase their predictive power. Furthermore, we address current limitations in assessment of deep-learning models, namely, comparison of single training run performance metrics, and offer a more robust solution.

SARS-CoV-2 Omicron variant spike RBD epitopes in complex with spike (S) -protein D614G mutations now predominate globally and increase infectivity by assembling more functional S proteins into the virion. On average, homeopathic medicine contains less than one molecule per dose on average. Single variable SphericalHarmonicY, LaguerreL, WhittakerM, SphericalBesselJ, LegendreP, LegendreQ, LaguerreL, ChebyshevT, and Hypergeometric1F1 Functions pFqarise in connection with the power series solution of the Schrodinger equation or in the summation of perturbation expansions in quantum mechanics. A quantum interpretation of the homeopathic method is presented here through a novel algebraic topology approach to supersymmetry and symmetry breaking in quantum fields and quantum gravity theory with the aim of developing a wide range of physical-based drug design applications in a Small Molecule Quantized Water Memory Network suggesting that the Hidden Black Hole Paradox of Information Loss might be solved under suitable conditions.