PathwayMap: Molecular Pathway Association with Self-Normalizing Neural Networks

Model organism life extending therapeutics modulate diverse nodes in the drug-gene-microbe tripartite human longevity interactome

Advances in antiaging drug/lead discovery in animal models constitute a large body of literature on novel senotherapeutics and geroprotectives. However, with little direct evidence or mechanism of action in humans-these drugs are utilized as nutraceuticals or repurposed supplements without proper testing directions, appropriate biomarkers, or consistent in-vivo models. In this study, we take previously identified drug candidates that have significant evidence of prolonging lifespan and promoting healthy aging in model organisms, and simulate them in human metabolic interactome networks. Screening for drug-likeness, toxicity, and KEGG network correlation scores, we generated a library of 285 safe and bioavailable compounds. We interrogated this library to present computational modeling-derived estimations of a tripartite interaction map of animal geroprotective compounds in the human molecular interactome extracted from longevity, senescence, and dietary restriction-associated genes. Our findings reflect previous studies in aging-associated metabolic disorders, and predict 25 best-connected drug interactors including Resveratrol, EGCG, Metformin, Trichostatin A, Caffeic Acid and Quercetin as direct modulators of lifespan and healthspan-associated pathways. We further clustered these compounds and the functionally enriched subnetworks therewith to identify longevity-exclusive, senescence-exclusive, pseudo-omniregulators and omniregulators within the set of interactome hub genes. Additionally, serum markers for drug-interactions, and interactions with potentially geroprotective gut microbial species distinguish the current study and present a holistic depiction of optimum gut microbial alteration by candidate drugs. These findings provide a systems level model of animal life-extending therapeutics in human systems, and act as precursors for expediting the ongoing global effort to find effective antiaging pharmacological interventions.Communicated by Ramaswamy H. Sarma.

Omics data analysis reveals common molecular basis of small cell lung cancer and COVID-19

The impact of COVID-19 infection on individuals with small cell lung cancer (SCLC) poses a serious threat. Unfortunately, the molecular basis of this severe comorbidity has yet to be elucidated. The present study addresses this gap utilizing publicly available omics data of COVID-19 and SCLC to explore the key molecules and associated pathways involved in the convergence of these diseases. Findings revealed 402 genes, that exhibited differential expression patterns in SCLC patients and also play a pivotal role in COVID-19 pathogenesis. Subsequent functional enrichment analyses identified relevant ontologies and pathways that are significantly associated with these genes, revealing important insights into their potential biological, molecular and cellular functions. The protein-protein interaction network, constructed under four combinatorial topological assessments, highlighted SMAD3, CAV1, PIK3R1, and FN1 as the primary components to this comorbidity. Our results suggest that these components significantly regulate this cross-talk triggering the PI3K-AKT and TGF-β signaling pathways. Lastly, this study made a multi-step computational attempt and identified corylifol A and ginkgetin from natural sources that can potentially inhibit these components. Therefore, the outcomes of this study offer novel perspectives on the common molecular mechanisms underlying SCLC and COVID-19 and present future opportunities for drug development.Communicated by Ramaswamy H. Sarma.

Next generation radiotheranostics promoting precision medicine

Radiotheranostics is a field of rapid growth with some approved treatments including 131I for thyroid cancer, 223Ra for osseous metastases, 177Lu-DOTATATE for neuroendocrine tumors, and 177Lu-PSMA (prostate-specific membrane antigen) for prostate cancer, and several more under investigation. In this review, we will cover the fundamentals of radiotheranostics, the key clinical studies that have led to current success, future developments with new targets, radionuclides and platforms, challenges with logistics and reimbursement and, lastly, forthcoming considerations regarding dosimetry, identifying the right line of therapy, artificial intelligence and more.

Artificial intelligence based virtual screening study for competitive and allosteric inhibitors of the SARS-CoV-2 main protease

SARS-CoV-2 is a highly contagious and dangerous coronavirus that first appeared in late 2019 causing COVID-19, a pandemic of acute respiratory illnesses that is still a threat to health and the general public safety. We performed deep docking studies of 800 M unique compounds in both the active and allosteric sites of the SARS-COV-2 Main Protease (Mpro) and the 15 M and 13 M virtual hits obtained were further taken for conventional docking and molecular dynamic (MD) studies. The best XP Glide docking scores obtained were -14.242 and -12.059 kcal/mol by CHEMBL591669 and the highest binding affinities were -10.5 kcal/mol (from 444215) and -11.2 kcal/mol (from NPC95421) for active and allosteric sites, respectively. Some hits can bind both sites making them a great area of concern. Re-docking of 8 random allosteric complexes in the active site shows a significant reduction in docking scores with a t-test P value of 2.532 × 10-11 at 95% confidence. Some specific interactions have higher elevations in docking scores. MD studies on 15 complexes show that single-ligand systems are stable as compared to double-ligand systems, and the allosteric binders identified are shown to modulate the active site binding as evidenced by the changes in the interaction patterns and stability of ligands and active site residues. When an allosteric complex was docked to the second monomer to check for homodimer formation, the validated homodimer could not be re-established, further supporting the potential of the identified allosteric binders. These findings could be important in developing novel therapeutics against SARS-CoV-2.Communicated by Ramaswamy H. Sarma.

Anti-Inflammatory and Anti-Diabetic Activity of Ferruginan, a Natural Compound from Olea ferruginea

Inflammation is a complex response of the human organism and relates to the onset of various disorders including diabetes. The current research work aimed at investigating the anti-inflammatory and anti-diabetic effects of ferruginan, a compound isolated from Olea ferruginea. Its in vitro anti-inflammatory activity was determined by using the heat-induced hemolysis assay, while the anti-diabetic effect of the compound was studied by the yeast cell glucose uptake assay. Ferruginan exhibited a maximum of 71.82% inhibition of inflammation and also increased the uptake of glucose by yeast cells by up to 74.96% at the highest tested concentration (100 µM). Moreover, ferruginan inhibited α-amylase dose-dependently, by up to 75.45% at the same concentration. These results indicated that ferruginan possesses promising anti-inflammatory and anti-diabetic properties in vitro, even if at high concentrations. To provide preliminary hypotheses on the potentially multi-target mechanisms underlying such effects, docking analyses were performed on α-amylase and on various molecular targets involved in inflammation such as 5′-adenosine monophosphate-activated protein kinase (AMPK, PDB ID 3AQV), cyclooxygenase (COX-1, PDB ID 1EQG, and COX-2, 1CX2), and tumor necrosis factor alpha (TNF-α, PDB ID 2AZ5). The docking studies suggested that the compound may act on α-amylase, COX-2, and AMPK.

Assessment of potassium ion channel during electric signalling in biofilm formation of Acinetobacter baumannii for finding antibiofilm molecule

Acinetobacter baumannii is an opportunistic ESKAPE pathogen which causes nosocomial infections and can produce biofilms that act as resistant determinants. The role of quorum sensing (chemical signaling) in biofilm establishment has already been studied extensively, but the existence of electrochemical signaling during biofilm formation by A. baumannii has not yet been investigated. The current study evaluated the presence of electrical signaling, types of ion channels involved, and their role in biofilm formation using spectroscopic and microbiological methods. The findings suggest that the potassium ion channel has a significant role in the electrical signaling during the biofilm formation by A. baumannii. Further, in-silico screening, molecular mechanics, and molecular dynamic simulation studies identify a potential lead, ZINC12496555(a specific inhibitor), which targets the potassium ion channel protein of A. baumannii. Mutational analysis of the interacting residues showed alterations in the unfolding rate of this protein after the selected mutation, which shows its role in the stability of this protein. It was also observed that identified lead has high antibiofilm activity, no human off-targets, and non-cytotoxicity to cell lines. Thus, identified lead against the potassium channel of A baumannii may be used as an effective therapeutic for the treatment of A. baumannii infections after further experimental validation.

Pharmacophore screening, denovo designing, retrosynthetic analysis, and combinatorial synthesis of a novel lead VTRA1.1 against RecA protein of Acinetobacter baumannii

Antibiotics and disinfectants resistance is acquired by activating RecA-mediated DNA repair, which maintains ROS-dependent DNA damage caused by the antimicrobial molecules. To increase the efficacy of different antimicrobials, an inhibitor can be developed against RecA protein. The present study aims to design a denovo inhibitor against RecA protein of Acinetobacter baumannii. Pharmacophore-based screening, molecular mechanics, molecular dynamics simulation (MDS), retrosynthetic analysis, and combinatorial synthesis were used to design lead VTRA1.1 against RecA of A. baumannii. Pharmacophore models (structure-based and ligand-based) were created, and a phase library of FDA-approved drugs was prepared. Screening of the phase library against these pharmacophore models selected thirteen lead molecules. These filtered leads were used for the denovo fragment-based design, which produced 253 combinations. These designed molecules were further analyzed for its interaction with active site of RecA that selected a hybrid VTRA1. Further, retrosynthetic analysis and combinatorial synthesis produced 1000 analogs of VTRA1 by more than 100 modifications. These analogs were used for XP docking, binding free energy calculation, and MDS analysis which finally select lead VTRA1.1 against RecA protein. Further, mutations at the interacting residues of RecA with VTRA1.1, alter the unfolding rate of RecA, which suggests the binding of VTRA1.1 to these residues may alter the stability of RecA. It is also found that VTRA1.1 had reduced interaction of RecA with LexA and ssDNA polydT, showing the lead's efficacy in controlling the SOS response. Further, it was also observed that VTRA1.1 does not contain any predicted human off-targets and no cytotoxicity to cell lines. As functional RecA is involved in antimicrobial resistance, denovo designed lead VTRA1.1 against RecA may be further developed as a significant combination for therapeutic uses against A. baumannii.

Complex machine learning model needs complex testing: Examining predictability of molecular binding affinity by a graph neural network

Drug discovery pipelines typically involve high-throughput screening of large amounts of compounds in a search of potential drugs candidates. As a chemical space of small organic molecules is huge, a "navigation" over it urges for fast and lightweight computational methods, thus promoting machine-learning approaches for processing huge pools of candidates. In this contribution, we present a graph-based deep neural network for prediction of protein-drug binding affinity and assess its predictive power under thorough testing conditions. Within the suggested approach, both protein and drug molecules are represented as graphs and passed to separate graph sub-networks, then concatenated and regressed towards a binding affinity. The neural network is trained on two binding affinity datasets-PDBbind and data imported from RCSB Protein Data Bank. In order to explore the generalization capabilities of the model we go beyond traditional random or leave-cluster-out techniques and demonstrate the need for more elaborate model performance assessment - six different strategies for test/train data partitioning (random, time- and property-arranged, protein- and ligand-clustered) with a k-fold cross-validation are engaged. Finally, we discuss the model performance in terms of a set of metrics for different split strategies and fold arrangement. Our code is available at https://github.com/SoftServeInc/affinity-by-GNN.

Therapeutic target database update 2022: facilitating drug discovery with enriched comparative data of targeted agents

Drug discovery relies on the knowledge of not only drugs and targets, but also the comparative agents and targets. These include poor binders and non-binders for developing discovery tools, prodrugs for improved therapeutics, co-targets of therapeutic targets for multi-target strategies and off-target investigations, and the collective structure-activity and drug-likeness landscapes of enhanced drug feature. However, such valuable data are inadequately covered by the available databases. In this study, a major update of the Therapeutic Target Database, previously featured in NAR, was therefore introduced. This update includes (a) 34 861 poor binders and 12 683 non-binders of 1308 targets; (b) 534 prodrug-drug pairs for 121 targets; (c) 1127 co-targets of 672 targets regulated by 642 approved and 624 clinical trial drugs; (d) the collective structure-activity landscapes of 427 262 active agents of 1565 targets; (e) the profiles of drug-like properties of 33 598 agents of 1102 targets. Moreover, a variety of additional data and function are provided, which include the cross-links to the target structure in PDB and AlphaFold, 159 and 1658 newly emerged targets and drugs, and the advanced search function for multi-entry target sequences or drug structures. The database is accessible without login requirement at: https://idrblab.org/ttd/.

Computer-Aided Design and Synthesis of a New Class of PEX14 Inhibitors: Substituted 2,3,4,5-Tetrahydrobenzo[F][1,4]oxazepines as Potential New Trypanocidal Agents

African and American trypanosomiases are estimated to affect several million people across the world, with effective treatments distinctly lacking. New, ideally oral, treatments with higher efficacy against these diseases are desperately needed. Peroxisomal import matrix (PEX) proteins represent a very interesting target for structure- and ligand-based drug design. The PEX5-PEX14 protein-protein interface in particular has been highlighted as a target, with inhibitors shown to disrupt essential cell processes in trypanosomes, leading to cell death. In this work, we present a drug development campaign that utilizes the synergy between structural biology, computer-aided drug design, and medicinal chemistry in the quest to discover and develop new potential compounds to treat trypanosomiasis by targeting the PEX14-PEX5 interaction. Using the structure of the known lead compounds discovered by Dawidowski et al. as the template for a chemically advanced template search (CATS) algorithm, we performed scaffold-hopping to obtain a new class of compounds with trypanocidal activity, based on 2,3,4,5-tetrahydrobenzo[f][1,4]oxazepines chemistry. The initial compounds obtained were taken forward to a first round of hit-to-lead optimization by synthesis of derivatives, which show activities in the range of low- to high-digit micromolar IC₅₀ in the in vitro tests. The NMR measurements confirm binding to PEX14 in solution, while immunofluorescent microscopy indicates disruption of protein import into the glycosomes, indicating that the PEX14-PEX5 protein-protein interface was successfully disrupted. These studies result in development of a novel scaffold for future lead optimization, while ADME testing gives an indication of further areas of improvement in the path from lead molecules toward a new drug active against trypanosomes.

In silico insights into potential gut microbial modulation of NAD+ metabolism and longevity

Recent evidence has prompted the notion of gut-microbial signatures as an indirect marker of aging and aging-associated decline in humans. However, the underlying host-symbiont molecular interactions contributing to these signatures remain poorly understood. In this study, we address this gap using cheminformatic analyses to elucidate potential gut microbial metabolites that may perturb the longevity-associated NAD+ metabolic network. In silico ADMET, KEGG interaction analysis, molecular docking, molecular dynamics simulation, and molecular mechanics calculation predict a large number of safe and bioavailable microbial metabolites to be direct and/or indirect activators of NAD+-dependent sirtuin proteins. Our simulation results suggest dihydropteroate, phenylpyruvic acid, indole-3-propionic acid, phenyllactic acid, all-trans-retinoic acid, and multiple deoxy-, methyl-, and cyclic nucleotides from intestinal microbiota as the best-performing regulators of NAD+ metabolism. Retracing these molecules to their source microorganisms also suggest commensal Escherichia, Bacteroides, Bifidobacteria, and Lactobacilli to be associated with the highest number of pro-longevity metabolites. These findings from our early-stage study, therefore, provide an informatics-based context for previous evidence in the area and grant novel insights for future clinical investigation intersecting anti-aging drug discovery, probiotics, and gut microbial signatures.

Coloring Molecules with Explainable Artificial Intelligence for Preclinical Relevance Assessment

Graph neural networks are able to solve certain drug discovery tasks such as molecular property prediction and de novo molecule generation. However, these models are considered "black-box" and "hard-to-debug". This study aimed to improve modeling transparency for rational molecular design by applying the integrated gradients explainable artificial intelligence (XAI) approach for graph neural network models. Models were trained for predicting plasma protein binding, hERG channel inhibition, passive permeability, and cytochrome P450 inhibition. The proposed methodology highlighted molecular features and structural elements that are in agreement with known pharmacophore motifs, correctly identified property cliffs, and provided insights into unspecific ligand-target interactions. The developed XAI approach is fully open-sourced and can be used by practitioners to train new models on other clinically relevant endpoints.

Marine phycocompound screening reveals a potential source of novel senotherapeutics

Cells undergo a controlled and systematic cycle of growth, replication and death. However, the integrity of this process gradually declines, leading to accumulation of senescent cells, a major hallmark of biological ageing. Dietary algae, particularly marine algae, have been long reported to exert anti-ageing benefits as cosmeceuticals and nutraceuticals with limited understanding of the molecular mechanisms underlying their activity. In this study, we have incorporated 1,202 previously reported bioactive small phycocompounds and subjected them to cheminformatic queries to assess these interactions. In-silico ADMET, 2-phase docking, metabolic pathway interaction and molecular dynamics simulations reveal multiple marine phycocompounds to have safe and effective senolytic potentials. We employed a novel deep convolutional neural network driven screening approach to identify (2R*, 3S*, 6R*, 7S*, 10R*, 13R*)-7,13-Dihydroxy-2,6-cyclo-1(9),14-xenicadiene-18,19-dial derived from Dilophus Fasciola, Laurendecumenyne A from Laurencia decumbens and 4-Bromo-3-ethyl-9-[(2E)-2-penten-4-yn-1-yl]-2,8-dioxabicyclo[5.2.1]decan-6-ol from Laurencia sp. to be potent inhibitors of multiple target senescent-cell anti-apoptotic pathway proteins. We simulated the best overall target inhibitors, specific protein inhibitors and molecular pathway regulators with each target protein and found stable interactions with minimum deviations (mean RMSD = 0.17 ± 0.01 nm) and gyrations (mean Rg = 1.64 ± 0.16 nm) of the simulated protein-compound complexes. Finally, molecular mechanics calculation suggests potent (mean ΔG = -69.56 ± 27.19 kCal/mol) and frequent hydrophobic interactions between the top performing marine phycocompounds and target proteins.

Design of drug-like hepsin inhibitors against prostate cancer and kidney stones

Hepsin, a transmembrane serine protease abundant in renal endothelial cells, is a promising therapeutic target against several cancers, particularly prostate cancer. It is involved in the release and polymerization of uromodulin in the urine, which plays a role in kidney stone formation. In this work, we design new potential hepsin inhibitors for high activity, improved specificity towards hepsin, and promising ADMET properties. The ligands were developed in silico through a novel hierarchical pipeline. This pipeline explicitly accounts for off-target binding to the related serine proteases matriptase and HGFA (human hepatocyte growth factor activator). We completed the pipeline incorporating ADMET properties of the candidate inhibitors into custom multi-objective optimization functions. The ligands designed show excellent prospects for targeting hepsin via the blood stream and the urine and thus enable key experimental studies. The computational pipeline proposed is remarkably cost-efficient and can be easily adapted for designing inhibitors against new drug targets.

A Deep-Learning Approach toward Rational Molecular Docking Protocol Selection

While a plethora of different protein–ligand docking protocols have been developed over the past twenty years, their performances greatly depend on the provided input protein–ligand pair. In this study, we developed a machine-learning model that uses a combination of convolutional and fully connected neural networks for the task of predicting the performance of several popular docking protocols given a protein structure and a small compound. We also rigorously evaluated the performance of our model using a widely available database of protein–ligand complexes and different types of data splits. We further open-source all code related to this study so that potential users can make informed selections on which protocol is best suited for their particular protein–ligand pair.