Structural basis for the severe adverse interaction of sofosbuvir and amiodarone on L-type Cav channels.

Drug-drug interaction of the antiviral sofosbuvir and the antiarrhythmics amiodarone has been reported to cause fatal heartbeat slowing. Sofosbuvir and its analog, MNI-1, were reported to potentiate the inhibition of cardiomyocyte calcium handling by amiodarone, which functions as a multi-channel antagonist, and implicate its inhibitory effect on L-type Cav channels, but the molecular mechanism has remained unclear. Here we present systematic cryo-EM structural analysis of Cav1.1 and Cav1.3 treated with amiodarone or sofosbuvir alone, or sofosbuvir/MNI-1 combined with amiodarone. Whereas amiodarone alone occupies the dihydropyridine binding site, sofosbuvir is not found in the channel when applied on its own. In the presence of amiodarone, sofosbuvir/MNI-1 is anchored in the central cavity of the pore domain through specific interaction with amiodarone and directly obstructs the ion permeation path. Our study reveals the molecular basis for the physical, pharmacodynamic interaction of two drugs on the scaffold of Cav channels.

Roles of Individual Human Cytochrome P450 Enzymes in Drug Metabolism.

Our knowledge of the roles of individual cytochrome P450 (P450, CYP) enzymes in drug metabolism has developed considerably in the past 30 years, and this base has been of considerable use in avoiding serious issues with drug interactions and issues due to variations. Some newer approaches are being considered for "phenotyping" of metabolism reactions with new drug candidates. Endogenous biomarkers are being used for non-invasive estimation of levels of individual P450 enzymes. There is also the matter of some remaining "orphan" P450s, which have yet to be assigned reactions. Practical problems that continue in drug development include predicting drug-drug interactions, predicting the effects of polymorphic and other P450 variations, and evaluating inter-species differences in drug metabolism, particularly in the context of "metabolism in safety testing" (MIST) regulatory issues ("disproportionate (human) metabolites"). Significance Statement Cytochrome P450 enzymes are the major catalysts involved in drug metabolism. The characterization of their individual roles has major implications in drug development and clinical practice.

Harnessing large language models' zero-shot and few-shot learning capabilities for regulatory research.

Large language models (LLMs) are sophisticated AI-driven models trained on vast sources of natural language data. They are adept at generating responses that closely mimic human conversational patterns. One of the most notable examples is OpenAI's ChatGPT, which has been extensively used across diverse sectors. Despite their flexibility, a significant challenge arises as most users must transmit their data to the servers of companies operating these models. Utilizing ChatGPT or similar models online may inadvertently expose sensitive information to the risk of data breaches. Therefore, implementing LLMs that are open source and smaller in scale within a secure local network becomes a crucial step for organizations where ensuring data privacy and protection has the highest priority, such as regulatory agencies. As a feasibility evaluation, we implemented a series of open-source LLMs within a regulatory agency's local network and assessed their performance on specific tasks involving extracting relevant clinical pharmacology information from regulatory drug labels. Our research shows that some models work well in the context of few- or zero-shot learning, achieving performance comparable, or even better than, neural network models that needed thousands of training samples. One of the models was selected to address a real-world issue of finding intrinsic factors that affect drugs' clinical exposure without any training or fine-tuning. In a dataset of over 700 000 sentences, the model showed a 78.5% accuracy rate. Our work pointed to the possibility of implementing open-source LLMs within a secure local network and using these models to perform various natural language processing tasks when large numbers of training examples are unavailable.

Wiring Between Close Nodes in Molecular Networks Evolves More Quickly Than Between Distant Nodes.

As species diverge, a wide range of evolutionary processes lead to changes in protein-protein interaction (PPI) networks and metabolic networks. The rate at which molecular networks evolve is an important question in evolutionary biology. Previous empirical work has focused on interactomes from model organisms to calculate rewiring rates, but this is limited by the relatively small number of species and sparse nature of network data across species. We present a proxy for variation in network topology: variation in drug-drug interactions (DDIs), obtained by studying drug combinations (DCs) across taxa. Here, we propose the rate at which DDIs change across species as an estimate of the rate at which the underlying molecular network changes as species diverge. We computed the evolutionary rates of DDIs using previously published data from a high-throughput study in gram-negative bacteria. Using phylogenetic comparative methods, we found that DDIs diverge rapidly over short evolutionary time periods, but that divergence saturates over longer time periods. In parallel, we mapped drugs with known targets in PPI and cofunctional networks. We found that the targets of synergistic DDIs are closer in these networks than other types of DCs and that synergistic interactions have a higher evolutionary rate, meaning that nodes that are closer evolve at a faster rate. Future studies of network evolution may use DC data to gain larger-scale perspectives on the details of network evolution within and between species.

Comprehensive evaluation of deep and graph learning on drug-drug interactions prediction.

Recent advances and achievements of artificial intelligence (AI) as well as deep and graph learning models have established their usefulness in biomedical applications, especially in drug-drug interactions (DDIs). DDIs refer to a change in the effect of one drug to the presence of another drug in the human body, which plays an essential role in drug discovery and clinical research. DDIs prediction through traditional clinical trials and experiments is an expensive and time-consuming process. To correctly apply the advanced AI and deep learning, the developer and user meet various challenges such as the availability and encoding of data resources, and the design of computational methods. This review summarizes chemical structure based, network based, natural language processing based and hybrid methods, providing an updated and accessible guide to the broad researchers and development community with different domain knowledge. We introduce widely used molecular representation and describe the theoretical frameworks of graph neural network models for representing molecular structures. We present the advantages and disadvantages of deep and graph learning methods by performing comparative experiments. We discuss the potential technical challenges and highlight future directions of deep and graph learning models for accelerating DDIs prediction.

DBGRU-SE: predicting drug-drug interactions based on double BiGRU and squeeze-and-excitation attention mechanism.

The prediction of drug-drug interactions (DDIs) is essential for the development and repositioning of new drugs. Meanwhile, they play a vital role in the fields of biopharmaceuticals, disease diagnosis and pharmacological treatment. This article proposes a new method called DBGRU-SE for predicting DDIs. Firstly, FP3 fingerprints, MACCS fingerprints, Pubchem fingerprints and 1D and 2D molecular descriptors are used to extract the feature information of the drugs. Secondly, Group Lasso is used to remove redundant features. Then, SMOTE-ENN is applied to balance the data to obtain the best feature vectors. Finally, the best feature vectors are fed into the classifier combining BiGRU and squeeze-and-excitation (SE) attention mechanisms to predict DDIs. After applying five-fold cross-validation, The ACC values of DBGRU-SE model on the two datasets are 97.51 and 94.98%, and the AUC are 99.60 and 98.85%, respectively. The results showed that DBGRU-SE had good predictive performance for drug-drug interactions.

Learning self-supervised molecular representations for drug-drug interaction prediction.

Drug-drug interactions (DDI) are a critical concern in healthcare due to their potential to cause adverse effects and compromise patient safety. Supervised machine learning models for DDI prediction need to be optimized to learn abstract, transferable features, and generalize to larger chemical spaces, primarily due to the scarcity of high-quality labeled DDI data. Inspired by recent advances in computer vision, we present SMR-DDI, a self-supervised framework that leverages contrastive learning to embed drugs into a scaffold-based feature space. Molecular scaffolds represent the core structural motifs that drive pharmacological activities, making them valuable for learning informative representations. Specifically, we pre-trained SMR-DDI on a large-scale unlabeled molecular dataset. We generated augmented views for each molecule via SMILES enumeration and optimized the embedding process through contrastive loss minimization between views. This enables the model to capture relevant and robust molecular features while reducing noise. We then transfer the learned representations for the downstream prediction of DDI. Experiments show that the new feature space has comparable expressivity to state-of-the-art molecular representations and achieved competitive DDI prediction results while training on less data. Additional investigations also revealed that pre-training on more extensive and diverse unlabeled molecular datasets improved the model's capability to embed molecules more effectively. Our results highlight contrastive learning as a promising approach for DDI prediction that can identify potentially hazardous drug combinations using only structural information.

Effectiveness of double-dose dolutegravir in people receiving rifampin-based tuberculosis treatment: an observational, cohort study of people with HIV from six countries.

BACKGROUND: Tenofovir-lamivudine-dolutegravir (TLD) is the preferred first-line antiretroviral therapy (ART) regimen. An additional 50 mg dose of dolutegravir (TLD + 50) is required with rifampin-containing tuberculosis (TB) co-treatment. There are limited data on the effectiveness of TLD + 50 in individuals with TB/HIV. METHODS: Prospective, observational cohort study at 12 sites in Haiti, Kenya, Malawi, South Africa, Uganda, Zimbabwe. Participants starting TLD and rifampin-containing TB treatment were eligible. Primary outcome was HIV-1 RNA ≤1000 copies/mL at end of TB treatment. FINDINGS: We enrolled 91 participants with TB/HIV: 75 (82%) ART-naïve participants starting TLD after a median 15 days on TB treatment, 10 (11%) ART-naïve participants starting TLD and TB treatment, 5 (5%) starting TB treatment after a median 3.3 years on TLD, and 1 (1%) starting TB treatment and TLD after changing from efavirenz/lamivudine/tenofovir. Median age was 37 years, 35% female, median CD4 count 120 cells/mm3 (IQR 50-295), 87% had HIV-1 RNA >1000 copies/mL. Two participants died during TB treatment. Among 89 surviving participants, 80 were followed to TB treatment completion, including 7 who had no HIV-1 RNA result due to missed visits. Primary virologic outcome was assessed in 73 participants, of whom 69 (95%, 95% CI 89-100%) had HIV-1 RNA ≤1000 copies/mL. No dolutegravir resistance mutations were detected among four participants with HIV-1 RNA >1000 copies/mL. INTERPRETATION: In routine programmatic settings, concurrent rifampin-containing TB treatment and TLD + 50 was feasible, well-tolerated, and achieved high rates of viral suppression in a cohort of predominantly ART-naïve people with TB/HIV.

Pharmacogenomics of Antiretroviral Drug Metabolism and Transport.

Antiretroviral therapy has markedly reduced morbidity and mortality for persons living with human immunodeficiency virus (HIV). Individual tailoring of antiretroviral regimens has the potential to further improve the long-term management of HIV through the mitigation of treatment failure and drug-induced toxicities. While the mechanisms underlying anti-HIV drug adverse outcomes are multifactorial, the application of drug-specific pharmacogenomic knowledge is required in order to move toward the personalization of HIV therapy. Thus, detailed understanding of the metabolism and transport of antiretrovirals and the influence of genetics on these pathways is important. To this end, this review provides an up-to-date overview of the metabolism of anti-HIV therapeutics and the impact of genetic variation in drug metabolism and transport on the treatment of HIV. Future perspectives on and current challenges in pursuing personalized HIV treatment are also discussed.

Real-world study: Impact of multidisciplinary management of apalutamide-associated adverse events in prostate cancer.

OBJECTIVE: To analyse the adverse events (AEs) associated with apalutamide and the impact of a multidisciplinary team (MDT) protocol on its management at a tertiary care hospital in a real-world setting. METHODS: This was an observational, prospective, cohort study based on real-world evidence at the Hospital Clínic de Barcelona. Includes patients diagnosed with metastatic hormone-sensitive prostate cancer (mHSPC) or high-risk nonmetastatic castration-resistant prostate cancer (nmCRPC) and who started treatment with apalutamide between May 2019 and March 2023 in a real-world clinical setting. RESULTS: Of the 121 patients treated with apalutamide, 52.1% experienced an AE, 19.8% experienced temporarily interruption or a reduction in the dose of apalutamide, and 13.2% discontinued treatment due to AEs. Without MDT protocol (49 patients), 24.5% of patients had to temporarily interrupt or reduce the dose of apalutamide due to AEs, with a median time from the start of treatment of 10.1 months, and 24.5% discontinued apalutamide due to AEs, with a median time from the start of treatment of 3.1 months. Meanwhile, whit MDT protocol (72 patients), 16.7% of patients had to temporarily interrupt or reduce the dose of apalutamide due to AEs, with a median time from the start of treatment of 1.6 months, and 5.6% discontinued apalutamide due to AEs, with a median time from the start of treatment of 4 months. The risk reduction associated with treatment discontinuation was statistically significant (p-value = 0.003). CONCLUSIONS: This study highlights the importance of MDT management of AEs associated with apalutamide to reduce treatment discontinuation.

Ivosidenib significantly reduces triazole levels in patients with acute myeloid leukemia and myelodysplastic syndrome.

BACKGROUND: Ivosidenib is primarily metabolized by CYP3A4; however, it induces CYP450 isozymes, including CYP3A4 and CYP2C9, whereas it inhibits drug transporters, including P-glycoprotein. Patients with acute myeloid leukemia are at risk of invasive fungal infections, and therefore posaconazole and voriconazole are commonly used in this population. Voriconazole is a substrate of CYP2C9, CYP2C19, and CYP3A4; therefore, concomitant ivosidenib may result in decreased serum concentrations. Although posaconazole is a substrate of P-glycoprotein, it is metabolized primarily via UDP glucuronidation; thus, the impact of ivosidenib on posaconazole exposure is unknown. METHODS: Patients treated with ivosidenib and concomitant triazole with at least one serum trough level were included. Subtherapeutic levels were defined as posaconazole <700 ng/mL and voriconazole <1.0 µg/mL. The incidences of breakthrough invasive fungal infections and QTc prolongation were identified at least 5 days after initiation of ivosidenib with concomitant triazole. RESULTS: Seventy-eight serum triazole levels from 31 patients receiving ivosidenib-containing therapy and concomitant triazole were evaluated. Of the 78 concomitant levels, 47 (60%) were subtherapeutic (posaconazole: n = 20 of 43 [47%]; voriconazole: n = 27 of 35 [77%]). Compared to levels drawn while patients were off ivosidenib, median triazole serum levels during concomitant ivosidenib were significantly reduced. There was no apparent increase in incidence of grade 3 QTc prolongation with concomitant azole antifungal and ivosidenib 500 mg daily. CONCLUSIONS: This study demonstrated that concomitant ivosidenib significantly reduced posaconazole and voriconazole levels. Voriconazole should be avoided, empiric high-dose posaconazole (>300 mg/day) may be considered, and therapeutic drug monitoring is recommended in all patients receiving concomitant ivosidenib.

Opioid metabolism and drug-drug interaction in cancer.

Concomitant use of multiple drugs in most patients with cancer may result in drug-drug interactions (DDIs), potentially causing serious adverse effects. These patients often experience unrelieved cancer-related pain (CRP) during and after cancer treatment, which can lead to a reduced quality of life. Opioids can be used as part of a multimodal pain management strategy when non-opioid analgesics are not providing adequate pain relief, not tolerated, or are contraindicated. However, due to their narrow therapeutic window, opioids are more susceptible to adverse events when a DDI occurs. Clinically relevant DDIs with opioids are usually pharmacokinetic, mainly occurring via metabolism by cytochrome P450 (CYP). This article aims to provide an overview of potential DDIs with opioids often used in the treatment of moderate-to-severe CRP and commonly used anticancer drugs such as chemotherapeutics, tyrosine kinase inhibitors (TKIs), or biologics. A DDI-checker tool was used to contextualize the tool-informed DDI assessment outcomes with clinical implications and practice. The findings were compared to observations from a literature search conducted in Embase and PubMed to identify clinical evidence for these potential DDIs. The limited results mainly included case studies and retrospective reviews. Some potential DDIs on the DDI-checker were aligned with literature findings, while others were contradictory. In conclusion, while DDI-checkers are useful tools in identifying potential DDIs, it is necessary to incorporate literature verification and comprehensive clinical assessment of the patient before implementing tool-informed decisions in clinical practice.

Effects of cannabidiol and Δ9-tetrahydrocannabinol on cytochrome P450 enzymes: a systematic review.

Due to legal, political, and cultural changes, the use of cannabis has rapidly increased in recent years. Research has demonstrated that the cannabinoids cannabidiol (CBD) and Δ9-tetrahydrocannabinol (THC) inhibit and induce cytochrome P450 (CYP450) enzymes. The objective of this review is to evaluate the effect of CBD and THC on the activity of CYP450 enzymes and the implications for drug-drug interactions (DDIs) with psychotropic agents that are CYP substrates. A systematic search was conducted using PubMed, Scopus, Scientific Electronic Library Online (SciELO) and PsychINFO. Search terms included 'cannabidiol', 'tetrahydrocannabinol', and 'cytochrome P450'. A total of seven studies evaluating the interaction of THC and CBD with CYP450 enzymes and psychotropic drugs were included. Both preclinical and clinical studies were included. Results from the included studies indicate that both CBD and THC inhibit several CYP450 enzymes including, but not limited to, CYP1A2, CYP3C19, and CYP2B6. While there are a few known CYP450 enzymes that are induced by THC and CBD, the induction of CYP450 enzymes is an understudied area of research and lacks clinical data. The inhibitory effects observed by CBD and THC on CYP450 enzymes vary in magnitude and may decrease the metabolism of psychotropic agents, cause changes in plasma levels of psychotropic medications, and increase adverse effects. Our findings clearly present interactions between THC and CBD and several CYP450 enzymes, providing clinicians evidence of a high risk of DDIs for patients who consume both cannabis and psychotropic medication. However, more clinical research is necessary before results are applied to clinical settings.

Insights into pharmacogenetics, drug-gene interactions, and drug-drug-gene interactions.

This review explores genetic contributors to drug interactions, known as drug-gene and drug-drug-gene interactions (DGI and DDGI, respectively). This article is part of a mini-review issue led by the International Society for the Study of Xenobiotics (ISSX) New Investigators Group. Pharmacogenetics (PGx) is the study of the impact of genetic variation on pharmacokinetics (PK), pharmacodynamics (PD), and adverse drug reactions. Genetic variation in pharmacogenes, including drug metabolizing enzymes and drug transporters, is common and can increase the risk of adverse drug events or contribute to reduced efficacy. In this review, we summarize clinically actionable genetic variants, and touch on methodologies such as genotyping patient DNA to identify genetic variation in targeted genes, and deep mutational scanning as a high-throughput in vitro approach to study the impact of genetic variation on protein function and/or expression in vitro. We highlight the utility of physiologically based pharmacokinetic (PBPK) models to integrate genetic and chemical inhibitor and inducer data for more accurate human PK simulations. Additionally, we analyze the limitations of historical ethnic descriptors in pharmacogenomics research. Altogether, the work herein underscores the importance of identifying and understanding complex DGI and DDGIs with the intention to provide better treatment outcomes for patients. We also highlight current barriers to wide-scale implementation of PGx-guided dosing as standard or care in clinical settings.

Transporter-mediated drug-drug interactions: regulatory guidelines, in vitro and in vivo methodologies and translation, special populations, and the blood-brain barrier.

This review, part of a special issue on drug-drug interactions (DDIs) spearheaded by the International Society for the Study of Xenobiotics (ISSX) New Investigators, explores the critical role of drug transporters in absorption, disposition, and clearance in the context of DDIs. Over the past two decades, significant advances have been made in understanding the clinical relevance of these transporters. Current knowledge on key uptake and efflux transporters that affect drug disposition and development is summarized. Regulatory guidelines from the FDA, EMA, and PMDA that inform the evaluation of potential transporter-mediated DDIs are discussed in detail. Methodologies for preclinical and clinical testing to assess potential DDIs are reviewed, with an emphasis on the utility of physiologically based pharmacokinetic (PBPK) modeling. This includes the application of relative abundance and expression factors to predict human pharmacokinetics (PK) using preclinical data, integrating the latest regulatory guidelines. Considerations for assessing transporter-mediated DDIs in special populations, including pediatric, hepatic, and renal impairment groups, are provided. Additionally, the impact of transporters at the blood-brain barrier (BBB) on the disposition of CNS-related drugs is explored. Enhancing the understanding of drug transporters and their role in drug disposition and toxicity can improve efficacy and reduce adverse effects. Continued research is essential to bridge remaining gaps in knowledge, particularly in comparison with cytochrome P450 (CYP) enzymes.

Prevalence and differences in the co-administration of drugs known to interact: an analysis of three distinct and large populations.

BACKGROUND: The co-administration of drugs known to interact greatly impacts morbidity, mortality, and health economics. This study aims to examine the drug-drug interaction (DDI) phenomenon with a large-scale longitudinal analysis of age and gender differences found in drug administration data from three distinct healthcare systems. METHODS: This study analyzes drug administrations from population-wide electronic health records in Blumenau (Brazil; 133 K individuals), Catalonia (Spain; 5.5 M individuals), and Indianapolis (USA; 264 K individuals). The stratified prevalences of DDI for multiple severity levels per patient gender and age at the time of administration are computed, and null models are used to estimate the expected impact of polypharmacy on DDI prevalence. Finally, to study actionable strategies to reduce DDI prevalence, alternative polypharmacy regimens using drugs with fewer known interactions are simulated. RESULTS: A large prevalence of co-administration of drugs known to interact is found in all populations, affecting 12.51%, 12.12%, and 10.06% of individuals in Blumenau, Indianapolis, and Catalonia, respectively. Despite very different healthcare systems and drug availability, the increasing prevalence of DDI as patients age is very similar across all three populations and is not explained solely by higher co-administration rates in the elderly. In general, the prevalence of DDI is significantly higher in women - with the exception of men over 50 years old in Indianapolis. Finally, we show that using proton pump inhibitor alternatives to omeprazole (the drug involved in more co-administrations in Catalonia and Blumenau), the proportion of patients that are administered known DDI can be reduced by up to 21% in both Blumenau and Catalonia and 2% in Indianapolis. CONCLUSIONS: DDI administration has a high incidence in society, regardless of geographic, population, and healthcare management differences. Although DDI prevalence increases with age, our analysis points to a complex phenomenon that is much more prevalent than expected, suggesting comorbidities as key drivers of the increase. Furthermore, the gender differences observed in most age groups across populations are concerning in regard to gender equity in healthcare. Finally, our study exemplifies how electronic health records' analysis can lead to actionable interventions that significantly reduce the administration of known DDI and its associated human and economic costs.

STNN-DDI: a Substructure-aware Tensor Neural Network to predict Drug-Drug Interactions.

Computational prediction of multiple-type drug-drug interaction (DDI) helps reduce unexpected side effects in poly-drug treatments. Although existing computational approaches achieve inspiring results, they ignore to study which local structures of drugs cause DDIs, and their interpretability is still weak. In this paper, by supposing that the interactions between two given drugs are caused by their local chemical structures (substructures) and their DDI types are determined by the linkages between different substructure sets, we design a novel Substructure-aware Tensor Neural Network model for DDI prediction (STNN-DDI). The proposed model learns a 3-D tensor of $\langle $  substructure, substructure, interaction type  $\rangle $ triplets, which characterizes a substructure-substructure interaction (SSI) space. According to a list of predefined substructures with specific chemical meanings, the mapping of drugs into this SSI space enables STNN-DDI to perform the multiple-type DDI prediction in both transductive and inductive scenarios in a unified form with an explicable manner. The comparison with deep learning-based state-of-the-art baselines demonstrates the superiority of STNN-DDI with the significant improvement of AUC, AUPR, Accuracy and Precision. More importantly, case studies illustrate its interpretability by both revealing an important substructure pair across drugs regarding a DDI type of interest and uncovering interaction type-specific substructure pairs in a given DDI. In summary, STNN-DDI provides an effective approach to predicting DDIs as well as explaining the interaction mechanisms among drugs. Source code is freely available at https://github.com/zsy-9/STNN-DDI.

Molormer: a lightweight self-attention-based method focused on spatial structure of molecular graph for drug-drug interactions prediction.

Multi-drug combinations for the treatment of complex diseases are gradually becoming an important treatment, and this type of treatment can take advantage of the synergistic effects among drugs. However, drug-drug interactions (DDIs) are not just all beneficial. Accurate and rapid identifications of the DDIs are essential to enhance the effectiveness of combination therapy and avoid unintended side effects. Traditional DDIs prediction methods use only drug sequence information or drug graph information, which ignores information about the position of atoms and edges in the spatial structure. In this paper, we propose Molormer, a method based on a lightweight attention mechanism for DDIs prediction. Molormer takes the two-dimension (2D) structures of drugs as input and encodes the molecular graph with spatial information. Besides, Molormer uses lightweight-based attention mechanism and self-attention distilling to process spatially the encoded molecular graph, which not only retains the multi-headed attention mechanism but also reduces the computational and storage costs. Finally, we use the Siamese network architecture to serve as the architecture of Molormer, which can make full use of the limited data to train the model for better performance and also limit the differences to some extent between networks dealing with drug features. Experiments show that our proposed method outperforms state-of-the-art methods in Accuracy, Precision, Recall and F1 on multi-label DDIs dataset. In the case study section, we used Molormer to make predictions of new interactions for the drugs Aliskiren, Selexipag and Vorapaxar and validated parts of the predictions. Code and models are available at https://github.com/IsXudongZhang/Molormer.

Attention-based Knowledge Graph Representation Learning for Predicting Drug-drug Interactions.

Drug-drug interactions (DDIs) are known as the main cause of life-threatening adverse events, and their identification is a key task in drug development. Existing computational algorithms mainly solve this problem by using advanced representation learning techniques. Though effective, few of them are capable of performing their tasks on biomedical knowledge graphs (KGs) that provide more detailed information about drug attributes and drug-related triple facts. In this work, an attention-based KG representation learning framework, namely DDKG, is proposed to fully utilize the information of KGs for improved performance of DDI prediction. In particular, DDKG first initializes the representations of drugs with their embeddings derived from drug attributes with an encoder-decoder layer, and then learns the representations of drugs by recursively propagating and aggregating first-order neighboring information along top-ranked network paths determined by neighboring node embeddings and triple facts. Last, DDKG estimates the probability of being interacting for pairwise drugs with their representations in an end-to-end manner. To evaluate the effectiveness of DDKG, extensive experiments have been conducted on two practical datasets with different sizes, and the results demonstrate that DDKG is superior to state-of-the-art algorithms on the DDI prediction task in terms of different evaluation metrics across all datasets.

A biomedical knowledge graph-based method for drug-drug interactions prediction through combining local and global features with deep neural networks.

Drug-drug interactions (DDIs) prediction is a challenging task in drug development and clinical application. Due to the extremely large complete set of all possible DDIs, computer-aided DDIs prediction methods are getting lots of attention in the pharmaceutical industry and academia. However, most existing computational methods only use single perspective information and few of them conduct the task based on the biomedical knowledge graph (BKG), which can provide more detailed and comprehensive drug lateral side information flow. To this end, a deep learning framework, namely DeepLGF, is proposed to fully exploit BKG fusing local-global information to improve the performance of DDIs prediction. More specifically, DeepLGF first obtains chemical local information on drug sequence semantics through a natural language processing algorithm. Then a model of BFGNN based on graph neural network is proposed to extract biological local information on drug through learning embedding vector from different biological functional spaces. The global feature information is extracted from the BKG by our knowledge graph embedding method. In DeepLGF, for fusing local-global features well, we designed four aggregating methods to explore the most suitable ones. Finally, the advanced fusing feature vectors are fed into deep neural network to train and predict. To evaluate the prediction performance of DeepLGF, we tested our method in three prediction tasks and compared it with state-of-the-art models. In addition, case studies of three cancer-related and COVID-19-related drugs further demonstrated DeepLGF's superior ability for potential DDIs prediction. The webserver of the DeepLGF predictor is freely available at http://120.77.11.78/DeepLGF/.