Synergy evaluation by a pathway-pathway interaction network: a new way to predict drug combination

Network-driven cancer cell avatars for combination discovery and biomarker identification for DNA damage response inhibitors

Combination therapy is well established as a key intervention strategy for cancer treatment, with the potential to overcome monotherapy resistance and deliver a more durable efficacy. However, given the scale of unexplored potential target space and the resulting combinatorial explosion, identifying efficacious drug combinations is a critical unmet need that is still evolving. In this paper, we demonstrate a network biology-driven, simulation-based solution, the Simulated Cell™. Integration of omics data with a curated signaling network enables the accurate and interpretable prediction of 66,348 combination-cell line pairs obtained from a large-scale combinatorial drug sensitivity screen of 684 combinations across 97 cancer cell lines (BAC = 0.62, AUC = 0.7). We highlight drug combination pairs that interact with DNA Damage Response pathways and are predicted to be synergistic, and deep network insight to identify biomarkers driving combination synergy. We demonstrate that the cancer cell 'avatars' capture the biological complexity of their in vitro counterparts, enabling the identification of pathway-level mechanisms of combination benefit to guide clinical translatability.

Harnessing machine learning to find synergistic combinations for FDA-approved cancer drugs

Combination therapy is a fundamental strategy in cancer chemotherapy. It involves administering two or more anti-cancer agents to increase efficacy and overcome multidrug resistance compared to monotherapy. However, drug combinations can exhibit synergy, additivity, or antagonism. This study presents a machine learning framework to classify and predict cancer drug combinations. The framework utilizes several key steps including data collection and annotation from the O'Neil drug interaction dataset, data preprocessing, stratified splitting into training and test sets, construction and evaluation of classification models to categorize combinations as synergistic, additive, or antagonistic, application of regression models to predict combination sensitivity scores for enhanced predictions compared to prior work, and the last step is examination of drug features and mechanisms of action to understand synergy behaviors for optimal combinations. The models identified combination pairs most likely to synergize against different cancers. Kinase inhibitors combined with mTOR inhibitors, DNA damage-inducing drugs or HDAC inhibitors showed benefit, particularly for ovarian, melanoma, prostate, lung and colorectal carcinomas. Analysis highlighted Gemcitabine, MK-8776 and AZD1775 as frequently synergizing across cancer types. This machine learning framework provides a valuable approach to uncover more effective multi-drug regimens.

Integrated transcriptome and cell phenotype analysis suggest involvement of PARP1 cleavage, Hippo/Wnt, TGF-β and MAPK signaling pathways in ovarian cancer cells response to cannabis and PARP1 inhibitor treatment

Introduction: Cannabis sativa is utilized mainly for palliative care worldwide. Ovarian cancer (OC) is a lethal gynecologic cancer. A particular cannabis extract fraction ('F7') and the Poly(ADP-Ribose) Polymerase 1 (PARP1) inhibitor niraparib act synergistically to promote OC cell apoptosis. Here we identified genetic pathways that are altered by the synergistic treatment in OC cell lines Caov3 and OVCAR3. Materials and methods: Gene expression profiles were determined by RNA sequencing and quantitative PCR. Microscopy was used to determine actin arrangement, a scratch assay to determine cell migration and flow cytometry to determine apoptosis, cell cycle and aldehyde dehydrogenase (ALDH) activity. Western blotting was used to determine protein levels. Results: Gene expression results suggested variations in gene expression between the two cell lines examined. Multiple genetic pathways, including Hippo/Wnt, TGF-β/Activin and MAPK were enriched with genes differentially expressed by niraparib and/or F7 treatments in both cell lines. Niraparib + F7 treatment led to cell cycle arrest and endoplasmic reticulum (ER) stress, inhibited cell migration, reduced the % of ALDH positive cells in the population and enhanced PARP1 cleavage. Conclusion: The synergistic effect of the niraparib + F7 may result from the treatment affecting multiple genetic pathways involving cell death and reducing mesenchymal characteristics.

Interpreting the Mechanism of Synergism for Drug Combinations Using Attention-Based Hierarchical Graph Pooling

Synergistic drug combinations provide huge potentials to enhance therapeutic efficacy and to reduce adverse reactions. However, effective and synergistic drug combination prediction remains an open question because of the unknown causal disease signaling pathways. Though various deep learning (AI) models have been proposed to quantitatively predict the synergism of drug combinations, the major limitation of existing deep learning methods is that they are inherently not interpretable, which makes the conclusions of AI models untransparent to human experts, henceforth limiting the robustness of the model conclusion and the implementation ability of these models in real-world human-AI healthcare. In this paper, we develop an interpretable graph neural network (GNN) that reveals the underlying essential therapeutic targets and the mechanism of the synergy (MoS) by mining the sub-molecular network of great importance. The key point of the interpretable GNN prediction model is a novel graph pooling layer, a self-attention-based node and edge pool (henceforth SANEpool), that can compute the attention score (importance) of genes and connections based on the genomic features and topology. As such, the proposed GNN model provides a systematic way to predict and interpret the drug combination synergism based on the detected crucial sub-molecular network. Experiments on various well-adopted drug-synergy-prediction datasets demonstrate that (1) the SANEpool model has superior predictive ability to generate accurate synergy score prediction, and (2) the sub-molecular networks detected by the SANEpool are self-explainable and salient for identifying synergistic drug combinations.

A systems level approach to study metabolic networks in prokaryotes with the aromatic amino acid biosynthesis pathway

Metabolism of an organism underlies its phenotype, which depends on many factors, such as the genetic makeup, habitat, and stresses to which it is exposed. This is particularly important for the prokaryotes, which undergo significant vertical and horizontal gene transfers. In this study we have used the energy-intensive Aromatic Amino Acid (Tryptophan, Tyrosine and Phenylalanine, TTP) biosynthesis pathway, in a large number of prokaryotes, as a model system to query the different levels of organization of metabolism in the whole intracellular biochemical network, and to understand how perturbations, such as mutations, affects the metabolic flux through the pathway - in isolation and in the context of other pathways connected to it. Using an agglomerative approach involving complex network analysis and Flux Balance Analyses (FBA), of the Tryptophan, Tyrosine and Phenylalanine and other pathways connected to it, we identify several novel results. Using the reaction network analysis and Flux Balance Analyses of the Tryptophan, Tyrosine and Phenylalanine and the genome-scale reconstructed metabolic pathways, many common hubs between the connected networks and the whole genome network are identified. The results show that the connected pathway network can act as a proxy for the whole genome network in Prokaryotes. This systems level analysis also points towards designing functional smaller synthetic pathways based on the reaction network and Flux Balance Analyses analysis.

PINet 1.0: A pathway network-based evaluation of drug combinations for the management of specific diseases

Drug combinations can increase the therapeutic effect by reducing the level of toxicity and the occurrence of drug resistance. Therefore, several drug combinations are often used in the management of complex diseases. However, due to the exponential growth in drug development, it would be impractical to evaluate all combinations through experiments. In view of this, we developed Pathway Interaction Network (PINet) biological model to estimate the optimal drug combinations for various diseases. The random walk with restart (RWR) algorithm was used to capture the “disease state” and “drug state,” while PINet was used to evaluate the optimal drug combinations and the high-order drug combination¹. The model achieved a mean area under the curve of a receiver operating characteristic curve of 0.885. In addition, for some diseases, PINet predicted the optimal drug combination. For example, in the case of acute myeloid leukemia, PINet correctly predicted midostaurin and gemtuzumab as effective drug combinations, as demonstrated by the results of a Phase-I clinical trial. Moreover, PINet also correctly predicted the potential drug combinations for diseases that lacked a training dataset that could not be predicted using standard machine learning models.

Transcriptome and metabolome analyses reveal the responses of brown planthoppers to RH resistant rice cultivar

The brown planthopper (BPH) Nilaparvata lugens (Stål) (Hemiptera: Delphacidae) is one of the most destructive rice pests in Asia. The application of insect-resistant rice cultivars is currently one of the principal means of controlling BPH. Understanding the physiological response mechanisms of BPH feeding on insect-resistant rice is the key for maintaining rice yield. Here, we measured the ecological fitness and analyzed the whole-body transcriptome and metabolome of BPH reared on susceptible cultivar Taichung Native 1 (TN1) and resistant cultivar Rathu Heenati (RH). Our results showed that RH significantly decreased the survival rate, female adult weight, honeydew secretion, the number of eggs laid per female and fat content of BPH. We identified 333 upregulated and 486 downregulated genes in BPH feeding on RH. These genes were mainly involved in energy metabolism, amino acid metabolism, hormone synthesis and vitamin metabolism pathways. We also detected 145 differentially accumulated metabolites in BPH reared on RH plants compared to BPH reared on TN1 plants, including multiple carbohydrates, amino acids, lipids, and some nucleosides. Combined analyses of transcriptome and metabolome showed that five pathways, including starch, sucrose, and galactose metabolism, were altered. The network for these pathways was subsequently visualized. Our results provide insights into the mechanisms of metabolite accumulation in BPH feeding on the RH rice variety. The results could help us better understand how insect-resistant rice cultivars combat BPH infestation, which is important for the comprehensive management of BPH.

Network Control Models With Personalized Genomics Data for Understanding Tumor Heterogeneity in Cancer

Due to rapid development of high-throughput sequencing and biotechnology, it has brought new opportunities and challenges in developing efficient computational methods for exploring personalized genomics data of cancer patients. Because of the high-dimension and small sample size characteristics of these personalized genomics data, it is difficult for excavating effective information by using traditional statistical methods. In the past few years, network control methods have been proposed to solve networked system with high-dimension and small sample size. Researchers have made progress in the design and optimization of network control principles. However, there are few studies comprehensively surveying network control methods to analyze the biomolecular network data of individual patients. To address this problem, here we comprehensively surveyed complex network control methods on personalized omics data for understanding tumor heterogeneity in precision medicine of individual patients with cancer.

DeSIDE-DDI: interpretable prediction of drug-drug interactions using drug-induced gene expressions

Adverse drug-drug interaction (DDI) is a major concern to polypharmacy due to its unexpected adverse side effects and must be identified at an early stage of drug discovery and development. Many computational methods have been proposed for this purpose, but most require specific types of information, or they have less concern in interpretation on underlying genes. We propose a deep learning-based framework for DDI prediction with drug-induced gene expression signatures so that the model can provide the expression level of interpretability for DDIs. The model engineers dynamic drug features using a gating mechanism that mimics the co-administration effects by imposing attention to genes. Also, each side-effect is projected into a latent space through translating embedding. As a result, the model achieved an AUC of 0.889 and an AUPR of 0.915 in unseen interaction prediction, which is competitively very accurate and outperforms other state-of-the-art methods. Furthermore, it can predict potential DDIs with new compounds not used in training. In conclusion, using drug-induced gene expression signatures followed by gating and translating embedding can increase DDI prediction accuracy while providing model interpretability. The source code is available on GitHub (https://github.com/GIST-CSBL/DeSIDE-DDI).

Machine Learning-Based Analysis of Glioma Grades Reveals Co-Enrichment

Gliomas develop and grow in the brain and central nervous system. Examining glioma grading processes is valuable for improving therapeutic challenges. One of the most extensive repositories storing transcriptomics data for gliomas is The Cancer Genome Atlas (TCGA). However, such big cohorts should be processed with caution and evaluated thoroughly as they can contain batch and other effects. Furthermore, biological mechanisms of cancer contain interactions among biomarkers. Thus, we applied an interpretable machine learning approach to discover such relationships. This type of transparent learning provides not only good predictability, but also reveals co-predictive mechanisms among features. In this study, we corrected the strong and confounded batch effect in the TCGA glioma data. We further used the corrected datasets to perform comprehensive machine learning analysis applied on single-sample gene set enrichment scores using collections from the Molecular Signature Database. Furthermore, using rule-based classifiers, we displayed networks of co-enrichment related to glioma grades. Moreover, we validated our results using the external glioma cohorts. We believe that utilizing corrected glioma cohorts from TCGA may improve the application and validation of any future studies. Finally, the co-enrichment and survival analysis provided detailed explanations for glioma progression and consequently, it should support the targeted treatment.

Detection of polypharmacy side effects by integrating multiple data sources and convolutional neural networks

The consumption of drug combinations, named polypharmacy, is commonly used for treating patients with several diseases or those with complex conditions. However, the main drawback of polypharmacy is the increased probability of harmful side effects. The polypharmacy side effects are caused by an interaction between two medications. It means that the drug-drug interaction causes changes in their activities due to interfering in each other's performance. Therefore, discovering these side effects is one of the most challenging and important aspects of drug production and consumption as it is associated with human health. In this paper, a method has been introduced for predicting the polypharmacy side effects, called PSECNN. It is a multi-label multi-class deep learning method that combines various basic features of drugs to predict the polypharmacy side effects. Firstly, PSECNN collects five basic features of drugs, such as individual drug's side effects, drug-protein interactions, chemical substructures, targets, and enzymes in order to create a novel combination of drug features. A feature extraction module creates five feature vectors with the same dimension for each drug based on the Jaccard similarity index. Based on the feature vectors, a unique representative is then created for each drug. These representative vectors are given in pairs as input to the deep neural network to predict the occurrence probability of side effects. According to the experimental evaluations, PSECNN could outperform the state-of-the-art polypharmacy side effects prediction methods up to 74%. It has been found that PSECNN has better performance with polypharmacy side effects with a cause of molecular basis due to the novel combination of basic drug features.

On the robustness of generalization of drug-drug interaction models

BACKGROUND

Deep learning methods are a proven commodity in many fields and endeavors. One of these endeavors is predicting the presence of adverse drug-drug interactions (DDIs). The models generated can predict, with reasonable accuracy, the phenotypes arising from the drug interactions using their molecular structures. Nevertheless, this task requires improvement to be truly useful. Given the complexity of the predictive task, an extensive benchmarking on structure-based models for DDIs prediction was performed to evaluate their drawbacks and advantages.

RESULTS

We rigorously tested various structure-based models that predict drug interactions using different splitting strategies to simulate different real-world scenarios. In addition to the effects of different training and testing setups on the robustness and generalizability of the models, we then explore the contribution of traditional approaches such as multitask learning and data augmentation.

CONCLUSION

Structure-based models tend to generalize poorly to unseen drugs despite their ability to identify new DDIs among drugs seen during training accurately. Indeed, they efficiently propagate information between known drugs and could be valuable for discovering new DDIs in a database. However, these models will most probably fail when exposed to unknown drugs. While multitask learning does not help in our case to solve the problem, the use of data augmentation does at least mitigate it. Therefore, researchers must be cautious of the bias of the random evaluation scheme, especially if their goal is to discover new DDIs.

A new humanized antibody is effective against pathogenic fungi in vitro

Invasive fungal infections mainly affect patients undergoing transplantation, surgery, neoplastic disease, immunocompromised subjects and premature infants, and cause over 1.5 million deaths every year. The most common fungi isolated in invasive diseases are Candida spp., Cryptococcus spp., and Aspergillus spp. and even if four classes of antifungals are available (Azoles, Echinocandins, Polyenes and Pyrimidine analogues), the side effects of drugs and fungal acquired and innate resistance represent the major hurdles to be overcome. Monoclonal antibodies are powerful tools currently used as diagnostic and therapeutic agents in different clinical contexts but not yet developed for the treatment of invasive fungal infections. In this paper we report the development of the first humanized monoclonal antibody specific for β-1,3 glucans, a vital component of several pathogenic fungi. H5K1 has been tested on C. auris, one of the most urgent threats and resulted efficient both alone and in combination with Caspofungin and Amphotericin B showing an enhancement effect. Our results support further preclinical and clinical developments for the use of H5K1 in the treatment of patients in need.

A neural network-based method for polypharmacy side effects prediction

BACKGROUND

Polypharmacy is a type of treatment that involves the concurrent use of multiple medications. Drugs may interact when they are used simultaneously. So, understanding and mitigating polypharmacy side effects are critical for patient safety and health. Since the known polypharmacy side effects are rare and they are not detected in clinical trials, computational methods are developed to model polypharmacy side effects.

RESULTS

We propose a neural network-based method for polypharmacy side effects prediction (NNPS) by using novel feature vectors based on mono side effects, and drug-protein interaction information. The proposed method is fast and efficient which allows the investigation of large numbers of polypharmacy side effects. Our novelty is defining new feature vectors for drugs and combining them with a neural network architecture to apply for the context of polypharmacy side effects prediction. We compare NNPS on a benchmark dataset to predict 964 polypharmacy side effects against 5 well-established methods and show that NNPS achieves better results than the results of all 5 methods in terms of accuracy, complexity, and running time speed. NNPS outperforms about 9.2% in Area Under the Receiver-Operating Characteristic, 12.8% in Area Under the Precision-Recall Curve, 8.6% in F-score, 10.3% in Accuracy, and 18.7% in Matthews Correlation Coefficient with 5-fold cross-validation against the best algorithm among other well-established methods (Decagon method). Also, the running time of the Decagon method which is 15 days for one fold of cross-validation is reduced to 8 h by the NNPS method.

CONCLUSIONS

The performance of NNPS is benchmarked against 5 well-known methods, Decagon, Concatenated drug features, Deep Walk, DEDICOM, and RESCAL, for 964 polypharmacy side effects. We adopt the 5-fold cross-validation for 50 iterations and use the average of the results to assess the performance of the NNPS method. The evaluation of the NNPS against five well-known methods, in terms of accuracy, complexity, and running time speed shows the performance of the presented method for an essential and challenging problem in pharmacology. Datasets and code for NNPS algorithm are freely accessible at https://github.com/raziyehmasumshah/NNPS .

The Synergistic Effects of 5-Aminosalicylic Acid and Vorinostat in the Treatment of Ulcerative Colitis

Background: The drug 5-aminosalicylic acid (5-ASA) is the first-line therapy for the treatment of patients with mild-to-moderate ulcerative colitis (UC). However, in some cases, 5-ASA cannot achieve the desired therapeutic effects. Therefore, patients have to undergo therapies that include corticosteroids, monoclonal antibodies or immunosuppressants, which are expensive and may be accompanied by significant side effects. Synergistic drug combinations can achieve greater therapeutic effects than individual drugs while contributing to combating drug resistance and lessening toxic side effects. Thus, in this study, we sought to identify synergistic drugs that can act synergistically with 5-ASA. Methods: We started our study with protein-metabolite analysis based on peroxisome proliferator-activated receptor gamma (PPARG), the therapeutic target of 5-ASA, to identify more additional potential drug targets. Then, we further evaluated the possibility of their synergy with PPARG by integrating Kyoto Encyclopedia of Genes and Genome (KEGG) pathway enrichment analysis, pathway-pathway interaction analysis, and semantic similarity analysis. Finally, we validated the synergistic effects with in vitro and in vivo experiments. Results: The combination of 5-ASA and vorinostat (SAHA) showed lower toxicity and mRNA expression of p65 in human colonic epithelial cell lines (Caco-2 and HCT-116), and more efficiently alleviated the symptoms of dextran sulfate sodium (DSS)-induced colitis than treatment with 5-ASA and SAHA alone. Conclusion: SAHA can exert effective synergistic effects with 5-ASA in the treatment of UC. One possible mechanism of synergism may be synergistic inhibition of the nuclear factor kappa B (NF-kB) signaling pathway. Moreover, the metabolite-butyric acid may be involved.

Machine Learning in Discovery of New Antivirals and Optimization of Viral Infections Therapy

Nowadays, computational approaches play an important role in the design of new drug-like compounds and optimization of pharmacotherapeutic treatment of diseases. The emerging growth of viral infections, including those caused by the Human Immunodeficiency Virus (HIV), Ebola virus, recently detected coronavirus, and some others, leads to many newly infected people with a high risk of death or severe complications. A huge amount of chemical, biological, clinical data is at the disposal of the researchers. Therefore, there are many opportunities to find the relationships between the particular features of chemical data and the antiviral activity of biologically active compounds based on machine learning approaches. Biological and clinical data can also be used for building models to predict relationships between viral genotype and drug resistance, which might help determine the clinical outcome of treatment. In the current study, we consider machine-learning approaches in the antiviral research carried out during the past decade. We overview in detail the application of machine-learning methods for the design of new potential antiviral agents and vaccines, drug resistance prediction, and analysis of virus-host interactions. Our review also covers the perspectives of using the machine-learning approaches for antiviral research, including Dengue, Ebola viruses, Influenza A, Human Immunodeficiency Virus, coronaviruses, and some others.

Rewired Pathways and Disrupted Pathway Crosstalk in Schizophrenia Transcriptomes by Multiple Differential Coexpression Methods

Transcriptomic studies of mental disorders using the human brain tissues have been limited, and gene expression signatures in schizophrenia (SCZ) remain elusive. In this study, we applied three differential co-expression methods to analyze five transcriptomic datasets (three RNA-Seq and two microarray datasets) derived from SCZ and matched normal postmortem brain samples. We aimed to uncover biological pathways where internal correlation structure was rewired or inter-coordination was disrupted in SCZ. In total, we identified 60 rewired pathways, many of which were related to neurotransmitter, synapse, immune, and cell adhesion. We found the hub genes, which were on the center of rewired pathways, were highly mutually consistent among the five datasets. The combinatory list of 92 hub genes was generally multi-functional, suggesting their complex and dynamic roles in SCZ pathophysiology. In our constructed pathway crosstalk network, we found “Clostridium neurotoxicity” and “signaling events mediated by focal adhesion kinase” had the highest interactions. We further identified disconnected gene links underlying the disrupted pathway crosstalk. Among them, four gene pairs (PAK1:SYT1, PAK1:RFC5, DCTN1:STX1A, and GRIA1:MAP2K4) were normally correlated in universal contexts. In summary, we systematically identified rewired pathways, disrupted pathway crosstalk circuits, and critical genes and gene links in schizophrenia transcriptomes.

Transcriptomic and Metabolomic Analyses of Diaphorina citri Kuwayama Infected and Non-infected With Candidatus Liberibacter Asiaticus

The Asian citrus psyllid Diaphorina citri is the transmission vector of Huanglongbing (HLB), a devastating disease of citrus plants. The bacterium “Candidatus Liberibacter asiaticus” (CLas) associated with HLB is transmitted between host plants by D. citri in a circulative manner. Understanding the interaction between CLas and its insect vector is key for protecting citrus cultivation from HLB damage. Here, we used RNA sequencing and liquid chromatography-mass spectrometry (LC-MS) to analyze the transcriptome and metabolome of D. citri interacting with CLas. We identified 662 upregulated and 532 downregulated genes in CLas-infected insects. These genes were enriched in pathways involving carbohydrate metabolism, the insects’ immune system, and metabolism of cofactors and vitamins. We also detected 105 differential metabolites between CLas-infected and non-infected insects, including multiple nucleosides and lipid-related molecules. The integrated analysis revealed nine pathways—including those of the glycine, serine, threonine, and purine metabolism—affected by the differentially expressed genes from both groups. The network for these pathways was subsequently constructed. Our results thus provide insights regarding the cross-talk between the transcriptomic and metabolomic changes in D. citri in response to CLas infection, as well as information on the pathways and genes/metabolites related to the CLas–D. citri interaction.

Protecting P-glycoprotein at the blood-brain barrier from degradation in an Alzheimer's disease mouse model

BACKGROUND

Failure to clear Aβ from the brain is partly responsible for Aβ brain accumulation in Alzheimer's disease (AD). A critical protein for clearing Aβ across the blood-brain barrier is the efflux transporter P-glycoprotein (P-gp). In AD, P-gp levels are reduced, which contributes to impaired Aβ brain clearance. However, the mechanism responsible for decreased P-gp levels is poorly understood and there are no strategies available to protect P-gp. We previously demonstrated in isolated brain capillaries ex vivo that human Aβ40 (hAβ40) triggers P-gp degradation by activating the ubiquitin-proteasome pathway. In this pathway, hAβ40 initiates P-gp ubiquitination, leading to internalization and proteasomal degradation of P-gp, which then results in decreased P-gp protein expression and transport activity levels. Here, we extend this line of research and present results from an in vivo study using a transgenic mouse model of AD (human amyloid precursor protein (hAPP)-overexpressing mice; Tg2576).

METHODS

In our study, hAPP mice were treated with vehicle, nocodazole (NCZ, microtubule inhibitor to block P-gp internalization), or a combination of NCZ and the P-gp inhibitor cyclosporin A (CSA). We determined P-gp protein expression and transport activity levels in isolated mouse brain capillaries and Aβ levels in plasma and brain tissue.

RESULTS

Treating hAPP mice with 5 mg/kg NCZ for 14 days increased P-gp levels to levels found in WT mice. Consistent with this, P-gp-mediated hAβ42 transport in brain capillaries was increased in NCZ-treated hAPP mice compared to untreated hAPP mice. Importantly, NCZ treatment significantly lowered hAβ40 and hAβ42 brain levels in hAPP mice, whereas hAβ40 and hAβ42 levels in plasma remained unchanged.

CONCLUSIONS

These findings provide in vivo evidence that microtubule inhibition maintains P-gp protein expression and transport activity levels, which in turn helps to lower hAβ brain levels in hAPP mice. Thus, protecting P-gp at the blood-brain barrier may provide a novel therapeutic strategy for AD and other Aβ-based pathologies.

Mapping drug-target interactions and synergy in multi-molecular therapeutics for pressure-overload cardiac hypertrophy

Advancements in systems biology have resulted in the development of network pharmacology, leading to a paradigm shift from "one-target, one-drug" to "target-network, multi-component therapeutics". We employ a chimeric approach involving in-vivo assays, gene expression analysis, cheminformatics, and network biology to deduce the regulatory actions of a multi-constituent Ayurvedic concoction, Amalaki Rasayana (AR) in animal models for its effect in pressure-overload cardiac hypertrophy. The proteomics analysis of in-vivo assays for Aorta Constricted and Biologically Aged rat models identify proteins expressed under each condition. Network analysis mapping protein-protein interactions and synergistic actions of AR using multi-component networks reveal drug targets such as ACADM, COX4I1, COX6B1, HBB, MYH14, and SLC25A4, as potential pharmacological co-targets for cardiac hypertrophy. Further, five out of eighteen AR constituents potentially target these proteins. We propose a distinct prospective strategy for the discovery of network pharmacological therapies and repositioning of existing drug molecules for treating pressure-overload cardiac hypertrophy.

Expression correlation attenuates within and between key signaling pathways in chronic kidney disease

Background

Compared to the conventional differential expression approach, differential coexpression analysis represents a different yet complementary perspective into diseased transcriptomes. In particular, global loss of transcriptome correlation was previously observed in aging mice, and a most recent study found genetic and environmental perturbations on human subjects tended to cause universal attenuation of transcriptome coherence. While methodological progresses surrounding differential coexpression have helped with research on several human diseases, there has not been an investigation of coexpression disruptions in chronic kidney disease (CKD) yet.

Methods

RNA-seq was performed on total RNAs of kidney tissue samples from 140 CKD patients. A combination of differential coexpression methods were employed to analyze the transcriptome transition in CKD from the early, mild phase to the late, severe kidney damage phase.

Results

We discovered a global expression correlation attenuation in CKD progression, with pathway Regulation of nuclear SMAD2/3 signaling demonstrating the most remarkable intra-pathway correlation rewiring. Moreover, the pathway Signaling events mediated by focal adhesion kinase displayed significantly weakened crosstalk with seven pathways, including Regulation of nuclear SMAD2/3 signaling. Well-known relevant genes, such as ACTN4, were characterized with widespread correlation disassociation with partners from a wide array of signaling pathways.

Conclusions

Altogether, our analysis reported a global expression correlation attenuation within and between key signaling pathways in chronic kidney disease, and presented a list of vanishing hub genes and disrupted correlations within and between key signaling pathways, illuminating on the pathophysiological mechanisms of CKD progression.

Predicting drug synergy for precision medicine using network biology and machine learning

Identification of effective drug combinations for patients is an expensive and time-consuming procedure, especially for in vitro experiments. To accelerate the synergistic drug discovery process, we present a new classification model to identify more effective anti-cancer drug pairs using in silico network biology approach. Based on the hypotheses that the drug synergy comes from the collective effects on the biological network, therefore, we developed six network biology features, including overlap and distance of drug perturbation network, that were derived by using individual drug-perturbed transcriptome profiles and the relevant biological network analysis. Using publicly available drug synergy databases and three machine-learning (ML) methods, the model was trained to discriminate the positive (synergistic) and negative (nonsynergistic) drug combinations. The proposed models were evaluated on the test cases to predict the most promising network biology feature, which is the network degree activity, i.e. the synergistic effect between drug pairs is mainly accounted by the complementary signaling pathways or molecular networks from two drugs.

Discovery of Synergistic and Antagonistic Drug Combinations against SARS-CoV-2 In Vitro

COVID-19 is undoubtedly the most impactful viral disease of the current century, afflicting millions worldwide. As yet, there is not an approved vaccine, as well as limited options from existing drugs for treating this disease. We hypothesized that combining drugs with independent mechanisms of action could result in synergy against SARS-CoV-2. Using in silico approaches, we prioritized 73 combinations of 32 drugs with potential activity against SARS-CoV-2 and then tested them in vitro . Overall, we identified 16 synergistic and 8 antagonistic combinations, 4 of which were both synergistic and antagonistic in a dose-dependent manner. Among the 16 synergistic cases, combinations of nitazoxanide with three other compounds (remdesivir, amodiaquine and umifenovir) were the most notable, all exhibiting significant synergy against SARS-CoV-2. The combination of nitazoxanide, an FDA-approved drug, and remdesivir, FDA emergency use authorization for the treatment of COVID-19, demonstrate a strong synergistic interaction. Notably, the combination of remdesivir and hydroxychloroquine demonstrated strong antagonism. Overall, our results emphasize the importance of both drug repurposing and preclinical testing of drug combinations for potential therapeutic use against SARS-CoV-2 infections.

Divergence and Convergence of Cerebral Ischemia Pathways Profile Deciphers Differential Pure Additive and Synergistic Mechanisms

Aim

The variable mechanisms on additive and synergistic effects of jasminoidin (JA)-Baicalin (BA) combination and JA-ursodeoxycholic acid (UA) combination in treating cerebral ischemia are not completely understood. In this study, we explored the differential pure mechanisms of additive and synergistic effects based on pathway analysis that excluded ineffective interference.

Methods

The MCAO mice were divided into eight groups: sham, vehicle, BA, JA, UA, Concha Margaritifera (CM), BA-JA combination (BJ), and JA-UA combination (JU). The additive and synergistic effects of combination groups were identified by cerebral infarct volume calculation. The differentially expressed genes based on a microarray chip containing 16,463 oligoclones were uploaded to GeneGo MetaCore software for pathway analyses and function catalogue. The comparison of specific pathways and functions crosstalk between different groups were analyzed to reveal the underlying additive and synergistic pharmacological variations.

Results

Additive BJ and synergistic JU were more effective than monotherapies of BA, JA, and UA, while CM was ineffective. Compared with monotherapies, 43 pathways and six functions were found uniquely in BJ group, with 33 pathways and three functions in JU group. We found six overlapping pathways and six overlapping functions between BJ and JU groups, which mainly involved central nervous system development. Thirty-seven specific pathways and 10 functions were activated by additive BJ, which were mainly related to cell adhesion and G-protein signaling; and 27 specific pathways and three functions of synergistic JU were associated with regulation of metabolism, DNA damage, and translation. The overlapping and distinct pathways and functions may contribute to different additive and synergistic effects.

Conclusion

The divergence pathways of pure additive effect of BJ were mainly related to cell adhesion and G-protein signaling, while the pure synergistic mechanism of JU depended on metabolism, translation and DNA damage. Such a systematic analysis of pathways may provide an important paradigm to reveal the pharmacological mechanisms underlying drug combinations.

Discovery of Novel Inhibitors Targeting Multi-UDP-hexose Pyrophosphorylases as Anticancer Agents

To minimize treatment toxicities, recent anti-cancer research efforts have switched from broad-based chemotherapy to targeted therapy, and emerging data show that altered cellular metabolism in cancerous cells can be exploited as new venues for targeted intervention. In this study, we focused on, among the altered metabolic processes in cancerous cells, altered glycosylation due to its documented roles in cancer tumorigenesis, metastasis and drug resistance. We hypothesize that the enzymes required for the biosynthesis of UDP-hexoses, glycosyl donors for glycan synthesis, could serve as therapeutic targets for cancers. Through structure-based virtual screening and kinetic assay, we identified a drug-like chemical fragment, GAL-012, that inhibit a small family of UDP-hexose pyrophosphorylases-galactose pyro-phosphorylase (GALT), UDP-glucose pyrophosphorylase (UGP2) and UDP-N-acetylglucosamine pyrophosphorylase (AGX1/UAP1) with an IC₅₀ of 30 µM. The computational docking studies supported the interaction of GAL-012 to the binding sites of GALT at Trp190 and Ser192, UGP2 at Gly116 and Lys127, and AGX1/UAP1 at Asn327 and Lys407, respectively. One of GAL-012 derivatives GAL-012-2 also demonstrated the inhibitory activity against GALT and UGP2. Moreover, we showed that GAL-012 suppressed the growth of PC3 cells in a dose-dependent manner with an EC₅₀ of 75 µM with no effects on normal skin fibroblasts at 200 µM. Western blot analysis revealed reduced expression of pAKT (Ser473), pAKT (Thr308) by 77% and 72%, respectively in the treated cells. siRNA experiments against the respective genes encoding the pyrophosphorylases were also performed and the results further validated the proposed roles in cancer growth inhibition. Finally, synergistic relationships between GAL-012 and tunicamycin, as well as bortezomib (BTZ) in killing cultured cancer cells were observed, respectively. With its unique scaffold and relatively small size, GAL-012 serves as a promising early chemotype for optimization to become a safe, effective, multi-target anti-cancer drug candidate which could be used alone or in combination with known therapeutics.

Machine Learning for Integrating Data in Biology and Medicine: Principles, Practice, and Opportunities

New technologies have enabled the investigation of biology and human health at an unprecedented scale and in multiple dimensions. These dimensions include myriad properties describing genome, epigenome, transcriptome, microbiome, phenotype, and lifestyle. No single data type, however, can capture the complexity of all the factors relevant to understanding a phenomenon such as a disease. Integrative methods that combine data from multiple technologies have thus emerged as critical statistical and computational approaches. The key challenge in developing such approaches is the identification of effective models to provide a comprehensive and relevant systems view. An ideal method can answer a biological or medical question, identifying important features and predicting outcomes, by harnessing heterogeneous data across several dimensions of biological variation. In this Review, we describe the principles of data integration and discuss current methods and available implementations. We provide examples of successful data integration in biology and medicine. Finally, we discuss current challenges in biomedical integrative methods and our perspective on the future development of the field.

Modeling polypharmacy side effects with graph convolutional networks

Motivation

The use of drug combinations, termed polypharmacy, is common to treat patients with complex diseases or co-existing conditions. However, a major consequence of polypharmacy is a much higher risk of adverse side effects for the patient. Polypharmacy side effects emerge because of drug-drug interactions, in which activity of one drug may change, favorably or unfavorably, if taken with another drug. The knowledge of drug interactions is often limited because these complex relationships are rare, and are usually not observed in relatively small clinical testing. Discovering polypharmacy side effects thus remains an important challenge with significant implications for patient mortality and morbidity.

Results

Here, we present Decagon, an approach for modeling polypharmacy side effects. The approach constructs a multimodal graph of protein-protein interactions, drug-protein target interactions and the polypharmacy side effects, which are represented as drug-drug interactions, where each side effect is an edge of a different type. Decagon is developed specifically to handle such multimodal graphs with a large number of edge types. Our approach develops a new graph convolutional neural network for multirelational link prediction in multimodal networks. Unlike approaches limited to predicting simple drug-drug interaction values, Decagon can predict the exact side effect, if any, through which a given drug combination manifests clinically. Decagon accurately predicts polypharmacy side effects, outperforming baselines by up to 69%. We find that it automatically learns representations of side effects indicative of co-occurrence of polypharmacy in patients. Furthermore, Decagon models particularly well polypharmacy side effects that have a strong molecular basis, while on predominantly non-molecular side effects, it achieves good performance because of effective sharing of model parameters across edge types. Decagon opens up opportunities to use large pharmacogenomic and patient population data to flag and prioritize polypharmacy side effects for follow-up analysis via formal pharmacological studies.

Availability and implementation

Source code and preprocessed datasets are at: http://snap.stanford.edu/decagon.

Natural Product Target Network Reveals Potential for Cancer Combination Therapies

A body of research demonstrates examples of in vitro and in vivo synergy between natural products and anti-neoplastic drugs for some cancers. However, the underlying biological mechanisms are still elusive. To better understand biological entities targeted by natural products and therefore provide rational evidence for future novel combination therapies for cancer treatment, we assess the targetable space of natural products using public domain compound-target information. When considering pathways from the Reactome database targeted by natural products, we found an increase in coverage of 61% (725 pathways), relative to pathways covered by FDA approved cancer drugs collected in the Cancer Targetome, a resource for evidence-based drug-target interactions. Not only is the coverage of pathways targeted by compounds increased when we include natural products, but coverage of targets within those pathways is also increased. Furthermore, we examined the distribution of cancer driver genes across pathways to assess relevance of natural products to critical cancer therapeutic space. We found 24 pathways enriched for cancer drivers that had no available cancer drug interactions at a potentially clinically relevant binding affinity threshold of < 100nM that had at least one natural product interaction at that same binding threshold. Assessment of network context highlighted the fact that natural products show target family groupings both distinct from and in common with cancer drugs, strengthening the complementary potential for natural products in the cancer therapeutic space. In conclusion, our study provides a foundation for developing novel cancer treatment with the combination of drugs and natural products.

Copolymeric micelles as efficient inert nanocarrier for hypericin in the photodynamic inactivation of Candida species

Aim: To evaluate the efficacy of photodynamic inactivation (PDI) mediated by hypericin encapsulated in P-123 copolymeric micelles (P123-Hyp) alone and in combination with fluconazole (FLU) against planktonic cells and biofilm formation of Candida species Materials & methods: PDI was performed using P123-Hyp and an LED device with irradiance of 3.0 mW/cm2 . Results: Most of isolates (70%) were completely inhibited with concentrations up to 2.0 μmol/l of HYP and light fluence of 16.2 J/cm2. FLU-resistant strains had synergic effect with P123-HYP-PDI and FLU. The biofilm formation was inhibited in all species, in additional the changes in Candida morphology observed by scanning electron microscopy. Conclusion: P123-Hyp-PDI is a promising option to treat fungal infections and medical devices to prevent biofilm formation and fungal spread.

A distinct strain of Arsenophonus symbiont decreases insecticide resistance in its insect host

Symbiotic bacteria are important drivers of phenotypic diversity in insects. One of the widespread symbionts to have emerged belongs to the genus Arsenophonus, however, its biological functions in most host insects remain entirely unknown. Here we report two distinct Arsenophonus strains in the brown planthopper (BPH), Nilaparvata lugens, a major pest insect in Asian countries that causes significant economic damage through rice crop destruction. Genomic resequencing data suggested that one Arsenophonus strain (S-type) negatively affected the insecticide resistance of the host. Indeed, replacement of the resident Arsenophonus with the S-type Arsenophonus significantly decreased host insecticide resistance. Transcriptome and metabolome analysis revealed down-regulation of xenobiotic metabolism and increased amino acid accumulation in the S-type Arsenophonus infected host. This study demonstrates how a symbiont-mediated phenotypic change can occur. The results of this study will aid in developing strategies that work through imposing an ecological disadvantage on insect pests, which will be of great value for pest control in agricultural industry.

The Emerging Potential for Network Analysis to Inform Precision Cancer Medicine

Precision cancer medicine promises to tailor clinical decisions to patients using genomic information. Indeed, successes of drugs targeting genetic alterations in tumors, such as imatinib that targets BCR-ABL in chronic myelogenous leukemia, have demonstrated the power of this approach. However, biological systems are complex, and patients may differ not only by the specific genetic alterations in their tumor, but also by more subtle interactions among such alterations. Systems biology and more specifically, network analysis, provides a framework for advancing precision medicine beyond clinical actionability of individual mutations. Here we discuss applications of network analysis to study tumor biology, early methods for N-of-1 tumor genome analysis, and the path for such tools to the clinic.

ChemDIS-Mixture: an online tool for analyzing potential interaction effects of chemical mixtures

The assessment of bioactivity and toxicity for mixtures remains a challenging work. Although several computational models have been developed to accelerate the evaluation of chemical-chemical interaction, a specific biological endpoint should be defined before applying the models that usually relies on clinical and experimental data. The development of computational methods is desirable for identifying potential biological endpoints of mixture interactions. To facilitate the identification of potential effects of mixture interactions, a novel online system named ChemDIS-Mixture is proposed to analyze the shared target proteins, and common enriched functions, pathways, and diseases affected by multiple chemicals. Venn diagram tools have been implemented for easy analysis and visualization of interaction targets and effects. Case studies have been provided to demonstrate the capability of ChemDIS-Mixture for identifying potential effects of mixture interactions in clinical studies. ChemDIS-Mixture provides useful functions for the identification of potential effects of coexposure to multiple chemicals. ChemDIS-Mixture is freely accessible at http://cwtung.kmu.edu.tw/chemdis/mixture .

In silico drug combination discovery for personalized cancer therapy

Background

Drug combination therapy, which is considered as an alternative to single drug therapy, can potentially reduce resistance and toxicity, and have synergistic efficacy. As drug combination therapies are widely used in the clinic for hypertension, asthma, and AIDS, they have also been proposed for the treatment of cancer. However, it is difficult to select and experimentally evaluate effective combinations because not only is the number of cancer drug combinations extremely large but also the effectiveness of drug combinations varies depending on the genetic variation of cancer patients. A computational approach that prioritizes the best drug combinations considering the genetic information of a cancer patient is necessary to reduce the search space.

Results

We propose an in-silico method for personalized drug combination therapy discovery. We predict the synergy between two drugs and a cell line using genomic information, targets of drugs, and pharmacological information. We calculate and predict the synergy scores of 583 drug combinations for 31 cancer cell lines. For feature dimension reduction, we select the mutations or expression levels of the genes in cancer-related pathways. We also used various machine learning models. Extremely Randomized Trees (ERT), a tree-based ensemble model, achieved the best performance in the synergy score prediction regression task. The correlation coefficient between the synergy scores predicted by ERT and the actual observations is 0.738. To compare with an existing drug combination synergy classification model, we reformulate the problem as a binary classification problem by thresholding the synergy scores. ERT achieved an F1 score of 0.954 when synergy scores of 20 and -20 were used as the threshold, which is 8.7% higher than that obtained by the state-of-the-art baseline model. Moreover, the model correctly predicts the most synergistic combination, from approximately 100 candidate drug combinations, as the top choice for 15 out of the 31 cell lines. For 28 out of the 31 cell lines, the model predicts the most synergistic combination in the top 10 of approximately 100 candidate drug combinations. Finally, we analyze the results, generate synergistic rules using the features, and validate the rules through the literature survey.

Conclusion

Using various types of genomic information of cancer cell lines, targets of drugs, and pharmacological information, a drug combination synergy prediction pipeline is proposed. The pipeline regresses the synergy level between two drugs and a cell line as well as classifies if there exists synergy or antagonism between them. Discovering new drug combinations by our pipeline may improve personalized cancer therapy.

Prediction of Effective Drug Combinations by an Improved Naïve Bayesian Algorithm

Drug combinatorial therapy is a promising strategy for combating complex diseases due to its fewer side effects, lower toxicity and better efficacy. However, it is not feasible to determine all the effective drug combinations in the vast space of possible combinations given the increasing number of approved drugs in the market, since the experimental methods for identification of effective drug combinations are both labor- and time-consuming. In this study, we conducted systematic analysis of various types of features to characterize pairs of drugs. These features included information about the targets of the drugs, the pathway in which the target protein of a drug was involved in, side effects of drugs, metabolic enzymes of the drugs, and drug transporters. The latter two features (metabolic enzymes and drug transporters) were related to the metabolism and transportation properties of drugs, which were not analyzed or used in previous studies. Then, we devised a novel improved naïve Bayesian algorithm to construct classification models to predict effective drug combinations by using the individual types of features mentioned above. Our results indicated that the performance of our proposed method was indeed better than the naïve Bayesian algorithm and other conventional classification algorithms such as support vector machine and K-nearest neighbor.

Ratiometric drug delivery using non-liposomal nanocarriers as an approach to increase efficacy and safety of combination chemotherapy

The observation that different drug ratios of the same drug combination can lead to synergistic or antagonistic effects when tested against the same cancer cell line in vitro gave rise to a new trend, the ratiometric delivery. This strategy consists of co-encapsulating a specific synergistic ratio of a drug combination into a nanocarrier so that synergism observed in vitro will be faithfully translated to in vivo, optimizing combination therapy. In this review we focus on how to quantify synergism in vitro, followed by how this affected the evolution of nanocarriers culminating in the ratiometric delivery, and finally we summarize the results of the non-liposomal formulations that were built upon this concept.

Biomolecular Network-Based Synergistic Drug Combination Discovery

Drug combination is a powerful and promising approach for complex disease therapy such as cancer and cardiovascular disease. However, the number of synergistic drug combinations approved by the Food and Drug Administration is very small. To bridge the gap between urgent need and low yield, researchers have constructed various models to identify synergistic drug combinations. Among these models, biomolecular network-based model is outstanding because of its ability to reflect and illustrate the relationships among drugs, disease-related genes, therapeutic targets, and disease-specific signaling pathways as a system. In this review, we analyzed and classified models for synergistic drug combination prediction in recent decade according to their respective algorithms. Besides, we collected useful resources including databases and analysis tools for synergistic drug combination prediction. It should provide a quick resource for computational biologists who work with network medicine or synergistic drug combination designing.

CD4-gp120 interaction interface - a gateway for HIV-1 infection in human: molecular network, modeling and docking studies

The major causative agent for Acquired Immune Deficiency Syndrome (AIDS) is Human Immunodeficiency Virus-1 (HIV-1). HIV-1 is a predominant subtype of HIV which counts on human cellular mechanism virtually in every aspect of its life cycle. Binding of viral envelope glycoprotein-gp120 with human cell surface CD4 receptor triggers the early infection stage of HIV-1. This study focuses on the interaction interface between these two proteins that play a crucial role for viral infectivity. The CD4-gp120 interaction interface has been studied through a comprehensive protein-protein interaction network (PPIN) analysis and highlighted as a useful step towards identifying potential therapeutic drug targets against HIV-1 infection. We prioritized gp41, Nef and Tat proteins of HIV-1 as valuable drug targets at early stage of viral infection. Lack of crystal structure has made it difficult to understand the biological implication of these proteins during disease progression. Here, computational protein modeling techniques and molecular dynamics simulations were performed to generate three-dimensional models of these targets. Besides, molecular docking was initiated to determine the desirability of these target proteins for already available HIV-1 specific drugs which indicates the usefulness of these protein structures to identify an effective drug combination therapy against AIDS.

Drug combinatorics and side effect estimation on the signed human drug-target network

Background

The mode of action of a drug on its targets can often be classified as being positive (activator, potentiator, agonist, etc.) or negative (inhibitor, blocker, antagonist, etc.). The signed edges of a drug-target network can be used to investigate the combined mechanisms of action of multiple drugs on the ensemble of common targets.

Results

In this paper it is shown that for the signed human drug-target network the majority of drug pairs tend to have synergistic effects on the common targets, i.e., drug pairs tend to have modes of action with the same sign on most of the shared targets, especially for the principal pharmacological targets of a drug. Methods are proposed to compute this synergism, as well as to estimate the influence of the drugs on the side effect of another drug.

Conclusions

Enriching a drug-target network with information of functional nature like the sign of the interactions allows to explore in a systematic way a series of network properties of key importance in the context of computational drug combinatorics.

Electronic supplementary material

The online version of this article (doi:10.1186/s12918-016-0326-8) contains supplementary material, which is available to authorized users.

NLLSS: Predicting Synergistic Drug Combinations Based on Semi-supervised Learning

Fungal infection has become one of the leading causes of hospital-acquired infections with high mortality rates. Furthermore, drug resistance is common for fungus-causing diseases. Synergistic drug combinations could provide an effective strategy to overcome drug resistance. Meanwhile, synergistic drug combinations can increase treatment efficacy and decrease drug dosage to avoid toxicity. Therefore, computational prediction of synergistic drug combinations for fungus-causing diseases becomes attractive. In this study, we proposed similar nature of drug combinations: principal drugs which obtain synergistic effect with similar adjuvant drugs are often similar and vice versa. Furthermore, we developed a novel algorithm termed Network-based Laplacian regularized Least Square Synergistic drug combination prediction (NLLSS) to predict potential synergistic drug combinations by integrating different kinds of information such as known synergistic drug combinations, drug-target interactions, and drug chemical structures. We applied NLLSS to predict antifungal synergistic drug combinations and showed that it achieved excellent performance both in terms of cross validation and independent prediction. Finally, we performed biological experiments for fungal pathogen Candida albicans to confirm 7 out of 13 predicted antifungal synergistic drug combinations. NLLSS provides an efficient strategy to identify potential synergistic antifungal combinations.

Drug combinations represent a promising strategy for overcoming fungal drug resistance and treating complex diseases. There is an urgent need to establish powerful computational methods for systematic prediction of synergistic drug combination on a large scale. Based on the assumption that principal drugs which obtain synergistic effect with similar adjuvant drugs are often similar and vice versa, NLLSS was developed to predict potential synergistic drug combinations by integrating known synergistic drug combinations, unlabeled drug combinations, drug-target interactions, and drug chemical structures. NLLSS has obtained the reliable performance in the cross validation and experimental validations, which indicated that NLLSS has an excellent performance of identifying potential synergistic drug combinations. Out of 13 predicted antifungal synergistic drug combinations, 7 candidates were experimentally confirmed. It is anticipated that NLLSS would be an important and useful resource by providing a new strategy to identify potential synergistic antifungal combinations, explore new indications of existing drugs, and provide useful insights into the underlying molecular mechanisms of synergistic drug combinations.