Computational drug repositioning with attention walking.

Drug repositioning aims to identify new therapeutic indications for approved medications. Recently, the importance of computational drug repositioning has been highlighted because it can reduce the costs, development time, and risks compared to traditional drug discovery. Most approaches in this area use networks for systematic analysis. Inferring drug-disease associations is then defined as a link prediction problem in a heterogeneous network composed of drugs and diseases. In this article, we present a novel method of computational drug repositioning, named drug repositioning with attention walking (DRAW). DRAW proceeds as follows: first, a subgraph enclosing the target link for prediction is extracted. Second, a graph convolutional network captures the structural features of the labeled nodes in the subgraph. Third, the transition probabilities are computed using attention mechanisms and converted into random walk profiles. Finally, a multi-layer perceptron takes random walk profiles and predicts whether a target link exists. As an experiment, we constructed two heterogeneous networks with drug-drug similarities based on chemical structures and anatomical therapeutic chemical classification (ATC) codes. Using 10-fold cross-validation, DRAW achieved an area under the receiver operating characteristic (ROC) curve of 0.903 and outperformed state-of-the-art methods. Moreover, we demonstrated the results of case studies for selected drugs and diseases to further confirm the capability of DRAW to predict drug-disease associations.

Elucidating the semantics-topology trade-off for knowledge inference-based pharmacological discovery.

Leveraging AI for synthesizing the deluge of biomedical knowledge has great potential for pharmacological discovery with applications including developing new therapeutics for untreated diseases and repurposing drugs as emergent pandemic treatments. Creating knowledge graph representations of interacting drugs, diseases, genes, and proteins enables discovery via embedding-based ML approaches and link prediction. Previously, it has been shown that these predictive methods are susceptible to biases from network structure, namely that they are driven not by discovering nuanced biological understanding of mechanisms, but based on high-degree hub nodes. In this work, we study the confounding effect of network topology on biological relation semantics by creating an experimental pipeline of knowledge graph semantic and topological perturbations. We show that the drop in drug repurposing performance from ablating meaningful semantics increases by 21% and 38% when mitigating topological bias in two networks. We demonstrate that new methods for representing knowledge and inferring new knowledge must be developed for making use of biomedical semantics for pharmacological innovation, and we suggest fruitful avenues for their development.

SGCLDGA: unveiling drug-gene associations through simple graph contrastive learning.

Drug repurposing offers a viable strategy for discovering new drugs and therapeutic targets through the analysis of drug-gene interactions. However, traditional experimental methods are plagued by their costliness and inefficiency. Despite graph convolutional network (GCN)-based models' state-of-the-art performance in prediction, their reliance on supervised learning makes them vulnerable to data sparsity, a common challenge in drug discovery, further complicating model development. In this study, we propose SGCLDGA, a novel computational model leveraging graph neural networks and contrastive learning to predict unknown drug-gene associations. SGCLDGA employs GCNs to extract vector representations of drugs and genes from the original bipartite graph. Subsequently, singular value decomposition (SVD) is employed to enhance the graph and generate multiple views. The model performs contrastive learning across these views, optimizing vector representations through a contrastive loss function to better distinguish positive and negative samples. The final step involves utilizing inner product calculations to determine association scores between drugs and genes. Experimental results on the DGIdb4.0 dataset demonstrate SGCLDGA's superior performance compared with six state-of-the-art methods. Ablation studies and case analyses validate the significance of contrastive learning and SVD, highlighting SGCLDGA's potential in discovering new drug-gene associations. The code and dataset for SGCLDGA are freely available at https://github.com/one-melon/SGCLDGA.

Neuroinflammation and Epilepsy: From Pathophysiology to Therapies Based on Repurposing Drugs.

Neuroinflammation and epilepsy are different pathologies, but, in some cases, they are so closely related that the activation of one of the pathologies leads to the development of the other. In this work, we discuss the three main cell types involved in neuroinflammation, namely (i) reactive astrocytes, (ii) activated microglia, and infiltration of (iii) peripheral immune cells in the central nervous system. Then, we discuss how neuroinflammation and epilepsy are interconnected and describe the use of different repurposing drugs with anti-inflammatory properties that have been shown to have a beneficial effect in different epilepsy models. This review reinforces the idea that compounds designed to alleviate seizures need to target not only the neuroinflammation caused by reactive astrocytes and microglia but also the interaction of these cells with infiltrated peripheral immune cells.

Drug Repositioning: A Monetary Stratagem to Discover a New Application of Drugs.

Drug repurposing, also referred to as drug repositioning or drug reprofiling, is a scientific approach to the detection of any new application for an already approved or investigational drug. It is a useful policy for the invention and development of new pharmacological or therapeutic applications of different drugs. The strategy has been known to offer numerous advantages over developing a completely novel drug for certain problems. Drug repurposing has numerous methodologies that can be categorized as target-oriented, drug-oriented, and problem-oriented. The choice of the methodology of drug repurposing relies on the accessible information about the drug molecule and like pharmacokinetic, pharmacological, physicochemical, and toxicological profile of the drug. In addition, molecular docking studies and other computer-aided methods have been known to show application in drug repurposing. The variation in dosage for original target diseases and novel diseases presents a challenge for researchers of drug repurposing in present times. The present review critically discusses the drugs repurposed for cancer, covid-19, Alzheimer's, and other diseases, strategies, and challenges of drug repurposing. Moreover, regulatory perspectives related to different countries like the United States (US), Europe, and India have been delineated in the present review.

Drug repurposing for cancer therapy.

Cancer, a complex and multifactorial disease, presents a significant challenge to global health. Despite significant advances in surgical, radiotherapeutic and immunological approaches, which have improved cancer treatment outcomes, drug therapy continues to serve as a key therapeutic strategy. However, the clinical efficacy of drug therapy is often constrained by drug resistance and severe toxic side effects, and thus there remains a critical need to develop novel cancer therapeutics. One promising strategy that has received widespread attention in recent years is drug repurposing: the identification of new applications for existing, clinically approved drugs. Drug repurposing possesses several inherent advantages in the context of cancer treatment since repurposed drugs are typically cost-effective, proven to be safe, and can significantly expedite the drug development process due to their already established safety profiles. In light of this, the present review offers a comprehensive overview of the various methods employed in drug repurposing, specifically focusing on the repurposing of drugs to treat cancer. We describe the antitumor properties of candidate drugs, and discuss in detail how they target both the hallmarks of cancer in tumor cells and the surrounding tumor microenvironment. In addition, we examine the innovative strategy of integrating drug repurposing with nanotechnology to enhance topical drug delivery. We also emphasize the critical role that repurposed drugs can play when used as part of a combination therapy regimen. To conclude, we outline the challenges associated with repurposing drugs and consider the future prospects of these repurposed drugs transitioning into clinical application.

Repurposing of Antifungal Drug Flucytosine/Flucytosine Cocrystals for Anticancer Activity against Prostate Cancer Targeting Apoptosis and Inflammatory Signaling Pathways.

This study aimed to repurpose the antifungal drug flucytosine (FCN) for anticancer activity together with cocrystals of nutraceutical coformers sinapic acid (SNP) and syringic acid (SYA). The cocrystal screening experiments with SNP resulted in three cocrystal hydrate forms in which two are polymorphs, namely, FCN-SNP F-I and FCN-SNP F-II, and the third one with different stoichiometry in the asymmetric unit (1:2:1 ratio of FCN:SNP:H2O, FCN-SNP F-III). Cocrystallization with SYA resulted in two hydrated cocrystal polymorphs, namely, FCN-SYA F-I and FCN-SYA F-II. All the cocrystal polymorphs were obtained concomitantly during the slow evaporation method, and one of the polymorphs of each system was produced in bulk by the slurry method. The interaction energy and lattice energies of all cocrystal polymorphs were established using solid-state DFT calculations, and the outcomes correlated with the experimental results. Further, the in vitro cytotoxic activity of the cocrystals was determined against DU145 prostate cancer and the results showed that the FCN-based cocrystals (FCN-SNP F-III and FCN-SYA F-I) have excellent growth inhibitory activity at lower concentrations compared with parent FCN molecules. The prepared cocrystals induce apoptosis by generating oxidative stress and causing nuclear damage in prostate cancer cells. The Western blot analysis also depicted that the cocrystals downregulate the inflammatory markers such as NLRP3 and caspase-1 and upregulate the intrinsic apoptosis signaling pathway marker proteins, such as Bax, p53, and caspase-3. These findings suggest that the antifungal drug FCN can be repurposed for anticancer activity.

Exploring the potential of drug repurposing for liver diseases: A comprehensive study.

Drug repurposing involves the investigation of existing drugs for new indications. It offers a great opportunity to quickly identify a new drug candidate at a lower cost than novel discovery and development. Despite the importance and potential role of drug repurposing, there is no specific definition that healthcare providers and the World Health Organization credit. Unfortunately, many similar and interchangeable concepts are being used in the literature, making it difficult to collect and analyze uniform data on repurposed drugs. This research was conducted based on understanding general criteria for drug repurposing, concentrating on liver diseases. Many drugs have been investigated for their effect on liver diseases even though they were originally approved (or on their way to being approved) for other diseases. Some of the hypotheses for drug repurposing were first captured from the literature and then processed further to test the hypothesis. Recently, with the revolution in bioinformatics techniques, scientists have started to use drug libraries and computer systems that can analyze hundreds of drugs to give a short list of candidates to be analyzed pharmacologically. However, this study revealed that drug repurposing is a potential aid that may help deal with liver diseases. It provides available or under-investigated drugs that could help treat hepatitis, liver cirrhosis, Wilson disease, liver cancer, and fatty liver. However, many further studies are needed to ensure the efficacy of these drugs on a large scale.

Dual Drug Repurposing: The Example of Saracatinib.

Saracatinib (AZD0530) is a dual Src/Abl inhibitor initially developed by AstraZeneca for cancer treatment; however, data from 2006 to 2024 reveal that this drug has been tested not only for cancer treatment, but also for the treatment of other diseases. Despite the promising pre-clinical results and the tolerability shown in phase I trials, where a maximum tolerated dose of 175 mg was defined, phase II clinical data demonstrated a low therapeutic action against several cancers and an elevated rate of adverse effects. Recently, pre-clinical research aimed at reducing the toxic effects and enhancing the therapeutic performance of saracatinib using nanoparticles and different pharmacological combinations has shown promising results. Concomitantly, saracatinib was repurposed to treat Alzheimer's disease, targeting Fyn. It showed great clinical results and required a lower daily dose than that defined for cancer treatment, 125 mg and 175 mg, respectively. In addition to Alzheimer's disease, this Src inhibitor has also been studied in relation to other health conditions such as pulmonary and liver fibrosis and even for analgesic and anti-allergic functions. Although saracatinib is still not approved by the Food and Drug Administration (FDA), the large number of alternative uses for saracatinib and the elevated number of pre-clinical and clinical trials performed suggest the huge potential of this drug for the treatment of different kinds of diseases.

Discovery of common molecular signatures and drug repurposing for COVID-19/Asthma comorbidity: ACE2 and multi-partite networks.

Angiotensin-converting enzyme 2 (ACE2) is identified as the functional receptor for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of the ongoing global coronavirus disease-2019 (COVID-19) pandemic. This study aimed to elucidate potential therapeutic avenues by scrutinizing approved drugs through the identification of the genetic signature associated with SARS-CoV-2 infection in individuals with asthma. This exploration was conducted through an integrated analysis, encompassing interaction networks between the ACE2 receptor and common host (co-host) factors implicated in COVID-19/asthma comorbidity. The comprehensive analysis involved the identification of common differentially expressed genes (cDEGs) and hub-cDEGs, functional annotations, interaction networks, gene set variation analysis (GSVA), gene set enrichment analysis (GSEA), and module construction. Interaction networks were used to identify overlapping disease modules and potential drug targets. Computational biology and molecular docking analyzes were utilized to discern functional drug modules. Subsequently, the impact of the identified drugs on the expression of hub-cDEGs was experimentally validated using a mouse model. A total of 153 cDEGs or co-host factors associated with ACE2 were identified in the COVID-19 and asthma comorbidity. Among these, seven significant cDEGs and proteins - namely, HRAS, IFNG, JUN, CDH1, TLR4, ICAM1, and SCD-were recognized as pivotal host factors linked to ACE2. Regulatory network analysis of hub-cDEGs revealed eight top-ranked transcription factors (TFs) proteins and nine microRNAs as key regulatory factors operating at the transcriptional and post-transcriptional levels, respectively. Molecular docking simulations led to the proposal of 10 top-ranked repurposable drug molecules (Rapamycin, Ivermectin, Everolimus, Quercetin, Estradiol, Entrectinib, Nilotinib, Conivaptan, Radotinib, and Venetoclax) as potential treatment options for COVID-19 in individuals with comorbid asthma. Validation analysis demonstrated that Rapamycin effectively inhibited ICAM1 expression in the HDM-stimulated mice group (p < 0.01). This study unveils the common pathogenesis and genetic signature underlying asthma and SARS-CoV-2 infection, delineated by the interaction networks of ACE2-related host factors. These findings provide valuable insights for the design and discovery of drugs aimed at more effective therapeutics within the context of lung disease comorbidities.

DeepSeq2Drug: An expandable ensemble end-to-end anti-viral drug repurposing benchmark framework by multi-modal embeddings and transfer learning.

Drug repurposing is promising in multiple scenarios, such as emerging viral outbreak controls and cost reductions of drug discovery. Traditional graph-based drug repurposing methods are limited to fast, large-scale virtual screens, as they constrain the counts for drugs and targets and fail to predict novel viruses or drugs. Moreover, though deep learning has been proposed for drug repurposing, only a few methods have been used, including a group of pre-trained deep learning models for embedding generation and transfer learning. Hence, we propose DeepSeq2Drug to tackle the shortcomings of previous methods. We leverage multi-modal embeddings and an ensemble strategy to complement the numbers of drugs and viruses and to guarantee the novel prediction. This framework (including the expanded version) involves four modal types: six NLP models, four CV models, four graph models, and two sequence models. In detail, we first make a pipeline and calculate the predictive performance of each pair of viral and drug embeddings. Then, we select the best embedding pairs and apply an ensemble strategy to conduct anti-viral drug repurposing. To validate the effect of the proposed ensemble model, a monkeypox virus (MPV) case study is conducted to reflect the potential predictive capability. This framework could be a benchmark method for further pre-trained deep learning optimization and anti-viral drug repurposing tasks. We also build software further to make the proposed model easier to reuse. The code and software are freely available at http://deepseq2drug.cs.cityu.edu.hk.

Comprehensive review of the repositioning of non-oncologic drugs for cancer immunotherapy.

Drug repositioning or repurposing has gained worldwide attention as a plausible way to search for novel molecules for the treatment of particular diseases or disorders. Drug repurposing essentially refers to uncovering approved or failed compounds for use in various diseases. Cancer is a deadly disease and leading cause of mortality. The search for approved non-oncologic drugs for cancer treatment involved in silico modeling, databases, and literature searches. In this review, we provide a concise account of the existing non-oncologic drug molecules and their therapeutic potential in chemotherapy. The mechanisms and modes of action of the repurposed drugs using computational techniques are also highlighted. Furthermore, we discuss potential targets, critical pathways, and highlight in detail the different challenges pertaining to drug repositioning for cancer immunotherapy.

A comparative benchmarking and evaluation framework for heterogeneous network-based drug repositioning methods.

Computational drug repositioning, which involves identifying new indications for existing drugs, is an increasingly attractive research area due to its advantages in reducing both overall cost and development time. As a result, a growing number of computational drug repositioning methods have emerged. Heterogeneous network-based drug repositioning methods have been shown to outperform other approaches. However, there is a dearth of systematic evaluation studies of these methods, encompassing performance, scalability and usability, as well as a standardized process for evaluating new methods. Additionally, previous studies have only compared several methods, with conflicting results. In this context, we conducted a systematic benchmarking study of 28 heterogeneous network-based drug repositioning methods on 11 existing datasets. We developed a comprehensive framework to evaluate their performance, scalability and usability. Our study revealed that methods such as HGIMC, ITRPCA and BNNR exhibit the best overall performance, as they rely on matrix completion or factorization. HINGRL, MLMC, ITRPCA and HGIMC demonstrate the best performance, while NMFDR, GROBMC and SCPMF display superior scalability. For usability, HGIMC, DRHGCN and BNNR are the top performers. Building on these findings, we developed an online tool called HN-DREP (http://hn-drep.lyhbio.com/) to facilitate researchers in viewing all the detailed evaluation results and selecting the appropriate method. HN-DREP also provides an external drug repositioning prediction service for a specific disease or drug by integrating predictions from all methods. Furthermore, we have released a Snakemake workflow named HN-DRES (https://github.com/lyhbio/HN-DRES) to facilitate benchmarking and support the extension of new methods into the field.

Single-cell transcriptional signature-based drug repurposing and in vitro evaluation in colorectal cancer.

BACKGROUND: The need for intelligent and effective treatment of diseases and the increase in drug design costs have raised drug repurposing as one of the effective strategies in biomedicine. There are various computational methods for drug repurposing, one of which is using transcription signatures, especially single-cell RNA sequencing (scRNA-seq) data, which show us a clear and comprehensive view of the inside of the cell to compare the state of disease and health. METHODS: In this study, we used 91,103 scRNA-seq samples from 29 patients with colorectal cancer (GSE144735 and GSE132465). First, differential gene expression (DGE) analysis was done using the ASAP website. Then we reached a list of drugs that can reverse the gene signature pattern from cancer to normal using the iLINCS website. Further, by searching various databases and articles, we found 12 drugs that have FDA approval, and so far, no one has reported them as a drug in the treatment of any cancer. Then, to evaluate the cytotoxicity and performance of these drugs, the MTT assay and real-time PCR were performed on two colorectal cancer cell lines (HT29 and HCT116). RESULTS: According to our approach, 12 drugs were suggested for the treatment of colorectal cancer. Four drugs were selected for biological evaluation. The results of the cytotoxicity analysis of these drugs are as follows: tezacaftor (IC10 = 19 µM for HCT-116 and IC10 = 2 µM for HT-29), fenticonazole (IC10 = 17 µM for HCT-116 and IC10 = 7 µM for HT-29), bempedoic acid (IC10 = 78 µM for HCT-116 and IC10 = 65 µM for HT-29), and famciclovir (IC10 = 422 µM for HCT-116 and IC10 = 959 µM for HT-29). CONCLUSIONS: Cost, time, and effectiveness are the main challenges in finding new drugs for diseases. Computational approaches such as transcriptional signature-based drug repurposing methods open new horizons to solve these challenges. In this study, tezacaftor, fenticonazole, and bempedoic acid can be introduced as promising drug candidates for the treatment of colorectal cancer. These drugs were evaluated in silico and in vitro, but it is necessary to evaluate them in vivo.

Multiscale topology in interactomic network: from transcriptome to antiaddiction drug repurposing.

The escalating drug addiction crisis in the United States underscores the urgent need for innovative therapeutic strategies. This study embarked on an innovative and rigorous strategy to unearth potential drug repurposing candidates for opioid and cocaine addiction treatment, bridging the gap between transcriptomic data analysis and drug discovery. We initiated our approach by conducting differential gene expression analysis on addiction-related transcriptomic data to identify key genes. We propose a novel topological differentiation to identify key genes from a protein-protein interaction network derived from DEGs. This method utilizes persistent Laplacians to accurately single out pivotal nodes within the network, conducting this analysis in a multiscale manner to ensure high reliability. Through rigorous literature validation, pathway analysis and data-availability scrutiny, we identified three pivotal molecular targets, mTOR, mGluR5 and NMDAR, for drug repurposing from DrugBank. We crafted machine learning models employing two natural language processing (NLP)-based embeddings and a traditional 2D fingerprint, which demonstrated robust predictive ability in gauging binding affinities of DrugBank compounds to selected targets. Furthermore, we elucidated the interactions of promising drugs with the targets and evaluated their drug-likeness. This study delineates a multi-faceted and comprehensive analytical framework, amalgamating bioinformatics, topological data analysis and machine learning, for drug repurposing in addiction treatment, setting the stage for subsequent experimental validation. The versatility of the methods we developed allows for applications across a range of diseases and transcriptomic datasets.

Methotrexate for Drug Repurposing as an Anti-Aggregatory Agent to Mercuric Treated α-Chymotrypsinogen-A.

Protein aggregation is related to numerous pathological conditions like Alzheimer's and Parkinson's disease. In our study, we have shown that an already existing FDA-approved drug; methotrexate (MTX) can be reprofiled on preformed α-chymotrypsinogen A (α-Cgn A) aggregates. The zymogen showed formation of aggregates upon interaction with mercuric ions, with increasing concentration of Hg2Cl2 (0-150 µM). The hike in ThT and ANS fluorescence concomitant with blue shift, bathochromic shift and the hyperchromic effect in the CR absorbance, RLS and turbidity measurements, substantiate the zymogen ß-rich aggregate formation. The secondary structural alterations of α- Cgn A as analyzed by CD measurements, FTIR and Raman spectra showed the transformation of native ß-barrel conformation to ß-inter-molecular rich aggregates. The native α- Cgn A have about 30% α-helical content which was found to be about 3% in presence of mercuric ions suggesting the formation of aggregates. The amorphous aggregates were visualized by SEM. On incubation of Hg2Cl2 treated α- Cgn A with increasing concentration of the MTX resulted in reversing aggregates to the native-like structure. These results were supported by remarkable decrease in ThT and ANS fluorescence intensities and CR absorbance and also consistent with CD, FTIR, and Raman spectroscopy data. MTX was found to increase the α-helical content of the zymogen from 3 to 15% proposing that drug is efficient in disrupting the ß-inter-molecular rich aggregates and reverting it to native like structure. The SEM images are in accordance with CD data showing the disintegration of aggregates. The most effective concentration of the drug was found to be 120 µM. Molecular docking analysis showed that MTX molecule was surrounded by the hydrophobic residues including Phe39, His40, Arg145, Tyr146, Thr151, Gly193, Ser195, and Gly216 and conventional hydrogen bonds, including Gln73 (bond length: 2.67Å), Gly142 (2.59Å), Thr144 (2.81Å), Asn150 (2.73Å), Asp153 (2.71Å), and Cys191 (2.53Å). This investigation will help to find the use of already existing drugs to cure protein misfolding-related abnormalities.

Artificial intelligence-driven drug repositioning uncovers efavirenz as a modulator of α-synuclein propagation: Implications in Parkinson's disease.

Parkinson's disease (PD) is a complex neurodegenerative disorder with an unclear etiology. Despite significant research efforts, developing disease-modifying treatments for PD remains a major unmet medical need. Notably, drug repositioning is becoming an increasingly attractive direction in drug discovery, and computational approaches offer a relatively quick and resource-saving method for identifying testable hypotheses that promote drug repositioning. We used an artificial intelligence (AI)-based drug repositioning strategy to screen an extensive compound library and identify potential therapeutic agents for PD. Our AI-driven analysis revealed that efavirenz and nevirapine, approved for treating human immunodeficiency virus infection, had distinct profiles, suggesting their potential effects on PD pathophysiology. Among these, efavirenz attenuated α-synuclein (α-syn) propagation and associated neuroinflammation in the brain of preformed α-syn fibrils-injected A53T α-syn Tg mice and α-syn propagation and associated behavioral changes in the C. elegans BiFC model. Through in-depth molecular investigations, we found that efavirenz can modulate cholesterol metabolism and mitigate α-syn propagation, a key pathological feature implicated in PD progression by regulating CYP46A1. This study opens new avenues for further investigation into the mechanisms underlying PD pathology and the exploration of additional drug candidates using advanced computational methodologies.

Signature reversion of three disease-associated gene signatures prioritizes cancer drug repurposing candidates.

Drug repurposing is promising because approving a drug for a new indication requires fewer resources than approving a new drug. Signature reversion detects drug perturbations most inversely related to the disease-associated gene signature to identify drugs that may reverse that signature. We assessed the performance and biological relevance of three approaches for constructing disease-associated gene signatures (i.e., limma, DESeq2, and MultiPLIER) and prioritized the resulting drug repurposing candidates for four low-survival human cancers. Our results were enriched for candidates that had been used in clinical trials or performed well in the PRISM drug screen. Additionally, we found that pamidronate and nimodipine, drugs predicted to be efficacious against the brain tumor glioblastoma (GBM), inhibited the growth of a GBM cell line and cells isolated from a patient-derived xenograft (PDX). Our results demonstrate that by applying multiple disease-associated gene signature methods, we prioritized several drug repurposing candidates for low-survival cancers.

EKGDR: An End-to-End Knowledge Graph-Based Method for Computational Drug Repurposing.

The lengthy and expensive process of developing new drugs from scratch, coupled with a high failure rate, has prompted the emergence of drug repurposing/repositioning as a more efficient and cost-effective approach. This approach involves identifying new therapeutic applications for existing approved drugs, leveraging the extensive drug-related data already gathered. However, the diversity and heterogeneity of data, along with the limited availability of known drug-disease interactions, pose significant challenges to computational drug design. To address these challenges, this study introduces EKGDR, an end-to-end knowledge graph-based approach for computational drug repurposing. EKGDR utilizes the power of a drug knowledge graph, a comprehensive repository of drug-related information that encompasses known drug interactions and various categorization information, as well as structural molecular descriptors of drugs. EKGDR employs graph neural networks, a cutting-edge graph representation learning technique, to embed the drug knowledge graph (nodes and relations) in an end-to-end manner. By doing so, EKGDR can effectively learn the underlying causes (intents) behind drug-disease interactions and recursively aggregate and combine relational messages between nodes along different multihop neighborhood paths (relational paths). This process generates representations of disease and drug nodes, enabling EKGDR to predict the interaction probability for each drug-disease pair in an end-to-end manner. The obtained results demonstrate that EKGDR outperforms previous models in all three evaluation metrics: area under the receiver operating characteristic curve (AUROC = 0.9475), area under the precision-recall curve (AUPRC = 0.9490), and recall at the top-200 recommendations (Recall@200 = 0.8315). To further validate EKGDR's effectiveness, we evaluated the top-20 candidate drugs suggested for each of Alzheimer's and Parkinson's diseases.

Utilizing machine learning to identify nifuroxazide as an inhibitor of ubiquitin-specific protease 21 in a drug repositioning strategy.

Ubiquitin-specific protease (USP), an enzyme catalyzing protein deubiquitination, is involved in biological processes related to metabolic disorders and cancer proliferation. We focused on constructing predictive models tailored to unveil compounds boasting USP21 inhibitory attributes. Six models, Extra Trees Classifier, Random Forest Classifier, LightGBM Classifier, XGBoost Classifier, Bagging Classifier, and a convolutional neural network harnessed from empirical data were selected for the screening process. These models guided our selection of 26 compounds from the FDA-approved drug library for further evaluation. Notably, nifuroxazide emerged as the most potent inhibitor, with a half-maximal inhibitory concentration of 14.9 ± 1.63 µM. The stability of protein-ligand complexes was confirmed using molecular modeling. Furthermore, nifuroxazide treatment of HepG2 cells not only inhibited USP21 and its established substrate ACLY but also elevated p-AMPKα, a downstream functional target of USP21. Intriguingly, we unveiled the previously unknown capacity of nifuroxazide to increase the levels of miR-4458, which was identified as downregulating USP21. This discovery was substantiated by manipulating miR-4458 levels in HepG2 cells, resulting in corresponding changes in USP21 protein levels in line with its predicted interaction with ACLY. Lastly, we confirmed the in vivo efficacy of nifuroxazide in inhibiting USP21 in mice livers, observing concurrent alterations in ACLY and p-AMPKα levels. Collectively, our study establishes nifuroxazide as a promising USP21 inhibitor with potential implications for addressing metabolic disorders and cancer proliferation. This multidimensional investigation sheds light on the intricate regulatory mechanisms involving USP21 and its downstream effects, paving the way for further exploration and therapeutic development.