This work leverages CODA TB, a groundbreaking dataset for a novel comprehensive method of early TB detection from medical big data. Departing from the erstwhile, we find mere cough duration less effective in TB prediction. We discover key demographic and clinical factors (e.g. heart rate, presenting symptoms) to be crucial in distinguishing TB cases, motivating comprehensive cough data analysis with enhanced screening.

Proteolysis-targeting chimeras (PROTACs) are hetero-bifunctional molecules that induce the degradation of target proteins by recruiting an E3 ligase. PROTACs have the potential to inactivate disease-related genes that are considered undruggable by small molecules, making them a promising therapy for the treatment of incurable diseases. However, only a few hundred proteins have been experimentally tested for their amenability to PROTACs, and it remains unclear which other proteins in the entire human genome can be targeted by PROTACs. In this study, we have developed PrePROTAC, an interpretable machine learning model based on a transformer-based protein sequence descriptor and random forest classification. PrePROTAC predicts genome-wide targets that can be degraded by CRBN, one of the E3 ligases. In the benchmark studies, PrePROTAC achieved a ROC-AUC of 0.81, an average precision of 0.84, and over 40% sensitivity at a false positive rate of 0.05. When evaluated by an external test set which comprised proteins from different structural folds than those in the training set, the performance of PrePROTAC did not drop significantly, indicating its generalizability. Furthermore, we developed an embedding SHapley Additive exPlanations (eSHAP) method, which extends conventional SHAP analysis for original features to an embedding space through in silico mutagenesis. This method allowed us to identify key residues in the protein structure that play critical roles in PROTAC activity. The identified key residues were consistent with existing knowledge. Using PrePROTAC, we identified over 600 novel understudied proteins that are potentially degradable by CRBN and proposed PROTAC compounds for three novel drug targets associated with Alzheimer's disease.

Systematically discovering protein-ligand interactions across the entire human and pathogen genomes is critical in chemical genomics, protein function prediction, drug discovery, and many other areas. However, more than 90% of gene families remain "dark"-i.e., their small-molecule ligands are undiscovered due to experimental limitations or human/historical biases. Existing computational approaches typically fail when the dark protein differs from those with known ligands. To address this challenge, we have developed a deep learning framework, called PortalCG, which consists of four novel components: (i) a 3-dimensional ligand binding site enhanced sequence pre-training strategy to encode the evolutionary links between ligand-binding sites across gene families; (ii) an end-to-end pretraining-fine-tuning strategy to reduce the impact of inaccuracy of predicted structures on function predictions by recognizing the sequence-structure-function paradigm; (iii) a new out-of-cluster meta-learning algorithm that extracts and accumulates information learned from predicting ligands of distinct gene families (meta-data) and applies the meta-data to a dark gene family; and (iv) a stress model selection step, using different gene families in the test data from those in the training and development data sets to facilitate model deployment in a real-world scenario. In extensive and rigorous benchmark experiments, PortalCG considerably outperformed state-of-the-art techniques of machine learning and protein-ligand docking when applied to dark gene families, and demonstrated its generalization power for target identifications and compound screenings under out-of-distribution (OOD) scenarios. Furthermore, in an external validation for the multi-target compound screening, the performance of PortalCG surpassed the rational design from medicinal chemists. Our results also suggest that a differentiable sequence-structure-function deep learning framework, where protein structural information serves as an intermediate layer, could be superior to conventional methodology where predicted protein structures were used for the compound screening. We applied PortalCG to two case studies to exemplify its potential in drug discovery: designing selective dual-antagonists of dopamine receptors for the treatment of opioid use disorder (OUD), and illuminating the understudied human genome for target diseases that do not yet have effective and safe therapeutics. Our results suggested that PortalCG is a viable solution to the OOD problem in exploring understudied regions of protein functional space.

Mental health is an ever-growing issue of concern, especially in light of the COVID pandemic. In this context, we study big data from social media over a 7-year time span to gauge evolving perceptions of mental health, and discuss our research findings, potentially useful for decision support in healthcare. We deploy topic modeling and sentiment analysis to estimate public perceptions of mental health issues, focusing on Twitter as the social media site. We claim that it is important to consider polarity as well as subjectivity in sentiment analysis to comprehend two different aspects of sentiment, i.e. orientation in the emotion, and extent of fact vs. opinion. We assert that ranking via topic modeling is beneficial to fathom the relative importance of issues over the years. We harness tools/techniques from natural language processing and data mining to discover knowledge from big data on social media, related to mental health. Some of our findings reveal that the sentiment around mental health has remained positive overall, but has decreased since the beginning of the COVID pandemic. Major events, such as elections and the pandemic, greatly impact the conversation surrounding mental health. Some topics have remained consistent throughout the years. In other topics, the tone of the public discussions has shifted. The outcomes of our study would be useful to a variety of professionals, ranging from data scientists to epidemiologists and psychologists. This work impacts big healthcare data in general.

molecular design is a key challenge in drug discovery due to the complexity of chemical space. With the availability of molecular datasets and advances in machine learning, many deep generative models are proposed for generating novel molecules with desired properties. However, most of the existing models focus only on molecular distribution learning and target-based molecular design, thereby hindering their potentials in real-world applications. In drug discovery, phenotypic molecular design has advantages over target-based molecular design, especially in first-in-class drug discovery. In this work, we propose the first deep graph generative model (FAME) targeting phenotypic molecular design, in particular gene expression-based molecular design. FAME leverages a conditional variational autoencoder framework to learn the conditional distribution generating molecules from gene expression profiles. However, this distribution is difficult to learn due to the complexity of the molecular space and the noisy phenomenon in gene expression data. To tackle these issues, a gene expression denoising (GED) model that employs contrastive objective function is first proposed to reduce noise from gene expression data. FAME is then designed to treat molecules as the sequences of fragments and learn to generate these fragments in autoregressive manner. By leveraging this fragment-based generation strategy and the denoised gene expression profiles, FAME can generate novel molecules with a high validity rate and desired biological activity. The experimental results show that FAME outperforms existing methods including both SMILES-based and graph-based deep generative models for phenotypic molecular design. Furthermore, the effective mechanism for reducing noise in gene expression data proposed in our study can be applied to omics data modeling in general for facilitating phenotypic drug discovery.

MOTIVATION: Drug discovery has witnessed intensive exploration of predictive modeling of drug-target physical interactions over two decades. However, a critical knowledge gap needs to be filled for correlating drug-target interactions with clinical outcomes: predicting genome-wide receptor activities or function selectivity, especially agonist versus antagonist, induced by novel chemicals. Two major obstacles compound the difficulty on this task: known data of receptor activity is far too scarce to train a robust model in light of genome-scale applications, and real-world applications need to deploy a model on data from various shifted distributions. RESULTS: To address these challenges, we have developed an end-to-end deep learning framework, DeepREAL, for multi-scale modeling of genome-wide ligand-induced receptor activities. DeepREAL utilizes self-supervised learning on tens of millions of protein sequences and pre-trained binary interaction classification to solve the data distribution shift and data scarcity problems. Extensive benchmark studies on G-protein coupled receptors (GPCRs), which simulate real-world scenarios, demonstrate that DeepREAL achieves state-of-the-art performances in out-of-distribution settings. DeepREAL can be extended to other gene families beyond GPCRs. AVAILABILITY AND IMPLEMENTATION: All data used are downloaded from Pfam (Mistry et al., 2020), GLASS (Chan et al., 2015) and IUPHAR/BPS and the data from reference (Sakamuru et al., 2021). Readers are directed to their official website for original data. Code is available on GitHub https://github.com/XieResearchGroup/DeepREAL. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Drug discovery is time-consuming and costly. Machine learning, especially deep learning, shows great potential in quantitative structure-activity relationship (QSAR) modeling to accelerate drug discovery process and reduce its cost. A big challenge in developing robust and generalizable deep learning models for QSAR is the lack of a large amount of data with high-quality and balanced labels. To address this challenge, we developed a self-training method, Partially LAbeled Noisy Student (PLANS), and a novel self-supervised graph embedding, Graph-Isomorphism-Network Fingerprint (GINFP), for chemical compounds representations with substructure information using unlabeled data. The representations can be used for predicting chemical properties such as binding affinity, toxicity, and others. PLANS-GINFP allows us to exploit millions of unlabeled chemical compounds as well as labeled and partially labeled pharmacological data to improve the generalizability of neural network models. RESULTS: We evaluated the performance of PLANS-GINFP for predicting Cytochrome P450 (CYP450) binding activity in a CYP450 dataset and chemical toxicity in the Tox21 dataset. The extensive benchmark studies demonstrated that PLANS-GINFP could significantly improve the performance in both cases by a large margin. Both PLANS-based self-training and GINFP-based self-supervised learning contribute to the performance improvement. CONCLUSION: To better exploit chemical structures as an input for machine learning algorithms, we proposed a self-supervised graph neural network-based embedding method that can encode substructure information. Furthermore, we developed a model agnostic self-training method, PLANS, that can be applied to any deep learning architectures to improve prediction accuracies. PLANS provided a way to better utilize partially labeled and unlabeled data. Comprehensive benchmark studies demonstrated their potentials in predicting drug metabolism and toxicity profiles using sparse, noisy, and imbalanced data. PLANS-GINFP could serve as a general solution to improve the predictive modeling for QSAR modeling.

BACKGROUND: Microbes are associated with many human diseases and influence drug efficacy. Small-molecule drugs may revolutionize biomedicine by fine-tuning the microbiota on the basis of individual patient microbiome signatures. However, emerging endeavors in small-molecule microbiome drug discovery continue to follow a conventional "one-drug-one-target-one-disease" process. A systematic pharmacology approach that would suppress multiple interacting pathogenic species in the microbiome, could offer an attractive alternative solution. RESULTS: We construct a disease-centric signed microbe-microbe interaction network using curated microbe metabolite information and their effects on host. We develop a Signed Random Walk with Restart algorithm for the accurate prediction of effect of microbes on human health and diseases. With a survey on the druggable and evolutionary space of microbe proteins, we find that 8-10% of them can be targeted by existing drugs or drug-like chemicals and that 25% of them have homologs to human proteins. We demonstrate that drugs for diabetes can be the lead compounds for development of microbiota-targeted therapeutics. We further show that the potential drug targets that specifically exist in pathogenic microbes are periplasmic and cellular outer membrane proteins. CONCLUSION: The systematic studies of the polypharmacological landscape of the microbiome network may open a new avenue for the small-molecule drug discovery of the microbiome. We believe that the application of systematic method on the polypharmacological investigation could lead to the discovery of novel drug therapies.

Predictive modeling of drug-induced gene expressions is a powerful tool for phenotype-based compound screening and drug repurposing. State-of-the-art machine learning methods use a small number of fixed cell lines as a surrogate for predicting actual expressions in a new cell type or tissue, although it is well known that drug responses depend on a cellular context. Thus, the existing approach has limitations when applied to personalized medicine, especially for many understudied diseases whose molecular profiles are dramatically different from those characterized in the training data. Besides the gene expression, dose-dependent cell viability is another important phenotype readout and is more informative than conventional summary statistics (e.g., IC50) for characterizing clinical drug efficacy and toxicity. However, few computational methods can reliably predict the dose-dependent cell viability. To address the challenges mentioned above, we designed a new deep learning model, MultiDCP, to predict cellular context-dependent gene expressions and cell viability on a specific dosage. The novelties of MultiDCP include a knowledge-driven gene expression profile transformer that enables context-specific phenotypic response predictions of novel cells or tissues, integration of multiple diverse labeled and unlabeled omics data, the joint training of the multiple prediction tasks, and a teacher-student training procedure that allows us to utilize unreliable data effectively. Comprehensive benchmark studies suggest that MultiDCP outperforms state-of-the-art methods with unseen cell lines that are dissimilar from the cell lines in the supervised training in terms of gene expressions. The predicted drug-induced gene expressions demonstrate a stronger predictive power than noisy experimental data for downstream tasks. Thus, MultiDCP is a useful tool for transcriptomics-based drug repurposing and compound screening that currently rely on noisy high-throughput experimental data. We applied MultiDCP to repurpose individualized drugs for Alzheimer's disease in terms of efficacy and toxicity, suggesting that MultiDCP is a potentially powerful tool for personalized drug discovery.

MOTIVATION: An unsolved fundamental problem in biology is to predict phenotypes from a new genotype under environmental perturbations. The emergence of multiple omics data provides new opportunities but imposes great challenges in the predictive modeling of genotype-phenotype associations. Firstly, the high-dimensionality of genomics data and the lack of coherent labeled data often make the existing supervised learning techniques less successful. Secondly, it is challenging to integrate heterogeneous omics data from different resources. Finally, few works have explicitly modeled the information transmission from DNA to phenotype, which involves multiple intermediate molecular types. Higher-level features (e.g. gene expression) usually have stronger discriminative and interpretable power than lower-level features (e.g. somatic mutation). RESULTS: We propose a novel Cross-LEvel Information Transmission (CLEIT) network framework to address the above issues. CLEIT aims to represent the asymmetrical multi-level organization of the biological system by integrating multiple incoherent omics data and to improve the prediction power of low-level features. CLEIT first learns the latent representation of the high-level domain then uses it as ground-truth embedding to improve the representation learning of the low-level domain in the form of contrastive loss. Besides, CLEIT can leverage the unlabeled heterogeneous omics data to improve the generalizability of the predictive model. We demonstrate the effectiveness and significant performance boost of CLEIT in predicting anti-cancer drug sensitivity from somatic mutations via the assistance of gene expressions when compared with state-of-the-art methods. CLEIT provides a general framework to model information transmissions and integrate multi-modal data in a multi-level system. AVAILABILITYAND IMPLEMENTATION: The source code is freely available at https://github.com/XieResearchGroup/CLEIT. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Drug combinations have demonstrated great potential in cancer treatments. They alleviate drug resistance and improve therapeutic efficacy. The fast-growing number of anti-cancer drugs has caused the experimental investigation of all drug combinations to become costly and time-consuming. Computational techniques can improve the efficiency of drug combination screening. Despite recent advances in applying machine learning to synergistic drug combination prediction, several challenges remain. First, the performance of existing methods is suboptimal. There is still much space for improvement. Second, biological knowledge has not been fully incorporated into the model. Finally, many models are lack interpretability, limiting their clinical applications. To address these challenges, we have developed a knowledge-enabled and self-attention transformer boosted deep learning model, TranSynergy, which improves the performance and interpretability of synergistic drug combination prediction. TranSynergy is designed so that the cellular effect of drug actions can be explicitly modeled through cell-line gene dependency, gene-gene interaction, and genome-wide drug-target interaction. A novel Shapley Additive Gene Set Enrichment Analysis (SA-GSEA) method has been developed to deconvolute genes that contribute to the synergistic drug combination and improve model interpretability. Extensive benchmark studies demonstrate that TranSynergy outperforms the state-of-the-art method, suggesting the potential of mechanism-driven machine learning. Novel pathways that are associated with the synergistic combinations are revealed and supported by experimental evidences. They may provide new insights into identifying biomarkers for precision medicine and discovering new anti-cancer therapies. Several new synergistic drug combinations have been predicted with high confidence for ovarian cancer which has few treatment options. The code is available at https://github.com/qiaoliuhub/drug_combination.

Identifying target genes of transcription factors (TFs) is crucial to understand transcriptional regulation. However, our understanding of genome-wide TF targeting profile is limited due to the cost of large-scale experiments and intrinsic complexity of gene regulation. Thus, computational prediction methods are useful to predict unobserved TF-gene associations. Here, we develop a new Weighted Imputed Neighborhood-regularized Tri-Factorization one-class collaborative filtering algorithm, WINTF. It predicts unobserved target genes for TFs using known but noisy, incomplete, and biased TF-gene associations and protein-protein interaction networks. Our benchmark study shows that WINTF significantly outperforms its counterpart matrix factorization-based algorithms and tri-factorization methods that do not include weight, imputation, and neighbor-regularization, for TF-gene association prediction. When evaluated by independent datasets, accuracy is 37.8 percent on the top 495 predicted associations, an enrichment factor of 4.19 compared with random guess. Furthermore, many predicted novel associations are supported by literature evidence. Although we only use canonical TF-gene interaction data, WINTF can directly be applied to tissue-specific data when available. Thus, WINTF provides a potentially useful framework to integrate multiple omics data for further improvement of TF-gene prediction and applications to other sparse and noisy biological data. The benchmark dataset and source code are freely available at https://github.com/XieResearchGroup/WINTF.

Graph Neural Networks (GNNs) are powerful for the representation learning of graph-structured data. Most of the GNNs use a message-passing scheme, where the embedding of a node is iteratively updated by aggregating the information from its neighbors. To achieve a better expressive capability of node influences, attention mechanism has grown to be popular to assign trainable weights to the nodes in aggregation. Though the attention-based GNNs have achieved remarkable results in various tasks, a clear understanding of their discriminative capacities is missing. In this work, we present a theoretical analysis of the representational properties of the GNN that adopts the attention mechanism as an aggregator. Our analysis determines all cases when those attention-based GNNs can always fail to distinguish certain distinct structures. Those cases appear due to the ignorance of cardinality information in attention-based aggregation. To improve the performance of attention-based GNNs, we propose cardinality preserved attention (CPA) models that can be applied to any kind of attention mechanisms. Our experiments on node and graph classification confirm our theoretical analysis and show the competitive performance of our CPA models. The code is available online: https://github.com/zetayue/CPA.

MOTIVATION: LINCS L1000 dataset contains numerous cellular expression data induced by large sets of perturbagens. Although it provides invaluable resources for drug discovery as well as understanding of disease mechanisms, the existing peak deconvolution algorithms cannot recover the accurate expression level of genes in many cases, inducing severe noise in the dataset and limiting its applications in biomedical studies. RESULTS: Here, we present a novel Bayesian-based peak deconvolution algorithm that gives unbiased likelihood estimations for peak locations and characterize the peaks with probability based z-scores. Based on the above algorithm, we build a pipeline to process raw data from L1000 assay into signatures that represent the features of perturbagen. The performance of the proposed pipeline is evaluated using similarity between the signatures of bio-replicates and the drugs with shared targets, and the results show that signatures derived from our pipeline gives a substantially more reliable and informative representation for perturbagens than existing methods. Thus, the new pipeline may significantly boost the performance of L1000 data in the downstream applications such as drug repurposing, disease modeling and gene function prediction. AVAILABILITY AND IMPLEMENTATION: The code and the precomputed data for LINCS L1000 Phase II (GSE 70138) are available at https://github.com/njpipeorgan/L1000-bayesian. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Computational prediction of a phenotypic response upon the chemical perturbation on a biological system plays an important role in drug discovery, and many other applications. Chemical fingerprints are a widely used feature to build machine learning models. However, the fingerprints that are derived from chemical structures ignore the biological context, thus, they suffer from several problems such as the activity cliff and curse of dimensionality. Fundamentally, the chemical modulation of biological activities is a multi-scale process. It is the genome-wide chemical-target interactions that modulate chemical phenotypic responses. Thus, the genome-scale chemical-target interaction profile will more directly correlate with in vitro and in vivo activities than the chemical structure. Nevertheless, the scope of direct application of the chemical-target interaction profile is limited due to the severe incompleteness, biasness, and noisiness of bioassay data. RESULTS: To address the aforementioned problems, we developed a novel chemical representation method: Latent Target Interaction Profile (LTIP). LTIP embeds chemicals into a low dimensional continuous latent space that represents genome-scale chemical-target interactions. Subsequently LTIP can be used as a feature to build machine learning models. Using the drug sensitivity of cancer cell lines as a benchmark, we have shown that the LTIP robustly outperforms chemical fingerprints regardless of machine learning algorithms. Moreover, the LTIP is complementary with the chemical fingerprints. It is possible for us to combine LTIP with other fingerprints to further improve the performance of bioactivity prediction. CONCLUSIONS: Our results demonstrate the potential of LTIP in particular and multi-scale modeling in general in predictive modeling of chemical modulation of biological activities.

Many complex diseases such as cancer are associated with multiple pathological manifestations. Moreover, the therapeutics for their treatments often lead to serious side effects. Thus, it is needed to develop multi-indication therapeutics that can simultaneously target multiple clinical indications of interest and mitigate the side effects. However, conventional one-drug-one-gene drug discovery paradigm and emerging polypharmacology approach rarely tackle the challenge of multi-indication drug design. For the first time, we propose a one-drug-multi-target-multi-indication strategy. We develop a novel structural systems pharmacology platform 3D-REMAP that uses ligand binding site comparison and protein-ligand docking to augment sparse chemical genomics data for the machine learning model of genome-scale chemical-protein interaction prediction. Experimentally validated predictions systematically show that 3D-REMAP outperforms state-of-the-art ligand-based, receptor-based, and machine learning methods alone. As a proof-of-concept, we utilize the concept of drug repurposing that is enabled by 3D-REMAP to design dual-indication anti-cancer therapy. The repurposed drug can demonstrate anti-cancer activity for cancers that do not have effective treatment as well as reduce the risk of heart failure that is associated with all types of existing anti-cancer therapies. We predict that levosimendan, a PDE inhibitor for heart failure, inhibits serine/threonine-protein kinase RIOK1 and other kinases. Subsequent experiments and systems biology analyses confirm this prediction, and suggest that levosimendan is active against multiple cancers, notably lymphoma, through the direct inhibition of RIOK1 and RNA processing pathway. We further develop machine learning models to predict cancer cell-line's and a patient's response to levosimendan. Our findings suggest that levosimendan can be a promising novel lead compound for the development of safe, effective, and precision multi-indication anti-cancer therapy. This study demonstrates the potential of structural systems pharmacology in designing polypharmacology for precision medicine. It may facilitate transforming the conventional one-drug-one-gene-one-disease drug discovery process and single-indication polypharmacology approach into a new one-drug-multi-target-multi-indication paradigm for complex diseases.

CRISPR-Cas is a powerful genome editing technology and has a great potential for in vivo gene therapy. Successful translational application of CRISPR-Cas to biomedicine still faces many safety concerns, including off-target side effect, cell fitness problem after CRISPR-Cas treatment, and on-target genome editing side effect in undesired tissues. To solve these issues, it is needed to design sgRNA with high cell-specific efficacy and specificity. Existing single-guide RNA (sgRNA) design tools mainly depend on a sgRNA sequence and the local information of the targeted genome, thus are not sufficient to account for the difference in the cellular response of the same gene in different cell types. To incorporate cell-specific information into the sgRNA design, we develop novel interpretable machine learning models, which integrate features learned from advanced transformer-based deep neural network with cell-specific gene property derived from biological network and gene expression profile, for the prediction of CRISPR-Cas9 and CRISPR-Cas12a efficacy and specificity. In benchmark studies, our models significantly outperform state-of-the-art algorithms. Furthermore, we find that the network-based gene property is critical for the prediction of cell-specific post-treatment cellular response. Our results suggest that the design of efficient and safe CRISPR-Cas needs to consider cell-specific information of genes. Our findings may bolster developing more accurate predictive models of CRISPR-Cas across a broad spectrum of biological conditions as well as provide new insight into developing efficient and safe CRISPR-based gene therapy.

Existing drug discovery processes follow a reductionist model of "one-drug-one-gene-one-disease," which is inadequate to tackle complex diseases involving multiple malfunctioned genes. The availability of big omics data offers opportunities to transform drug discovery process into a new paradigm of systems pharmacology that focuses on designing drugs to target molecular interaction networks instead of a single gene. Here, we develop a reliable multi-rank, multi-layered recommender system, ANTENNA, to mine large-scale chemical genomics and disease association data for prediction of novel drug-gene-disease associations. ANTENNA integrates a novel tri-factorization based dual-regularized weighted and imputed One Class Collaborative Filtering (OCCF) algorithm, tREMAP, with a statistical framework based on Random Walk with Restart and assess the reliability of specific predictions. In the benchmark, tREMAP clearly outperforms the single-rank OCCF. We apply ANTENNA to a real-world problem: repurposing old drugs for new clinical indications without effective treatments. We discover that FDA-approved drug diazoxide can inhibit multiple kinase genes responsible for many diseases including cancer and kill triple negative breast cancer (TNBC) cells efficiently [Formula: see text]. TNBC is a deadly disease without effective targeted therapies. Our finding demonstrates the power of big data analytics in drug discovery and developing a targeted therapy for TNBC.

It is a grand challenge to reveal the causal effects of DNA variants in complex phenotypes. Although statistical techniques can establish correlations between genotypes and phenotypes in Genome-Wide Association Studies (GWAS), they often fail when the variant is rare. The emerging Network-based Association Studies aim to address this shortcoming in statistical analysis, but are mainly applied to coding variations. Increasing evidences suggest that non-coding variants play critical roles in the etiology of complex diseases. However, few computational tools are available to study the effect of rare non-coding variants on phenotypes. Here we have developed a multiscale modeling variant-to-function-to-network framework VariFunNet to address these challenges. VariFunNet first predict the functional variations of molecular interactions, which result from the non-coding variants. Then we incorporate the genes associated with the functional variation into a tissue-specific gene network, and identify subnetworks that transmit the functional variation to molecular phenotypes. Finally, we quantify the functional implication of the subnetwork, and prioritize the association of the non-coding variants with the phenotype. We have applied VariFunNet to investigating the causal effect of rare non-coding variants on Alzheimer's disease (AD). Among top 21 ranked causal non-coding variants, 16 of them are directly supported by existing evidences. The remaining 5 novel variants dysregulate multiple downstream biological processes, all of which are associated with the pathology of AD. Furthermore, we propose potential new drug targets that may modulate diverse pathways responsible for AD. These findings may shed new light on discovering new biomarkers and therapies for the prevention, diagnosis, and treatment of AD. Our results suggest that multiscale modeling is a potentially powerful approach to studying causal genotype-phenotype associations.

MOTIVATION: A complex disease is usually driven by a number of genes interwoven into networks, rather than a single gene product. Network comparison or differential network analysis has become an important means of revealing the underlying mechanism of pathogenesis and identifying clinical biomarkers for disease classification. Most studies, however, are limited to network correlations that mainly capture the linear relationship among genes, or rely on the assumption of a parametric probability distribution of gene measurements. They are restrictive in real application. RESULTS: We propose a new Joint density based non-parametric Differential Interaction Network Analysis and Classification (JDINAC) method to identify differential interaction patterns of network activation between two groups. At the same time, JDINAC uses the network biomarkers to build a classification model. The novelty of JDINAC lies in its potential to capture non-linear relations between molecular interactions using high-dimensional sparse data as well as to adjust confounding factors, without the need of the assumption of a parametric probability distribution of gene measurements. Simulation studies demonstrate that JDINAC provides more accurate differential network estimation and lower classification error than that achieved by other state-of-the-art methods. We apply JDINAC to a Breast Invasive Carcinoma dataset, which includes 114 patients who have both tumor and matched normal samples. The hub genes and differential interaction patterns identified were consistent with existing experimental studies. Furthermore, JDINAC discriminated the tumor and normal sample with high accuracy by virtue of the identified biomarkers. JDINAC provides a general framework for feature selection and classification using high-dimensional sparse omics data. AVAILABILITY AND IMPLEMENTATION: R scripts available at https://github.com/jijiadong/JDINAC. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The increasingly connected world has catalyzed the fusion of networks from different domains, which facilitates the emergence of a new network model-multi-layered networks. Examples of such kind of network systems include critical infrastructure networks, biological systems, organization-level collaborations, cross-platform e-commerce, and so forth. One crucial structure that distances multi-layered network from other network models is its cross-layer dependency, which describes the associations between the nodes from different layers. Needless to say, the cross-layer dependency in the network plays an essential role in many data mining applications like system robustness analysis and complex network control. However, it remains a daunting task to know the exact dependency relationships due to noise, limited accessibility, and so forth. In this article, we tackle the cross-layer dependency inference problem by modeling it as a collective collaborative filtering problem. Based on this idea, we propose an effective algorithm Fascinate that can reveal unobserved dependencies with linear complexity. Moreover, we derive Fascinate-ZERO, an online variant of Fascinate that can respond to a newly added node timely by checking its neighborhood dependencies. We perform extensive evaluations on real datasets to substantiate the superiority of our proposed approaches.

Multi-layered networks have recently emerged as a new network model, which naturally finds itself in many high-impact application domains, ranging from critical inter-dependent infrastructure networks, biological systems, organization-level collaborations, to cross-platform e-commerce, etc. Cross-layer dependency, which describes the dependencies or the associations between nodes across different layers/networks, often plays a central role in many data mining tasks on such multi-layered networks. Yet, it remains a daunting task to accurately know the cross-layer dependency a prior. In this paper, we address the problem of inferring the missing cross-layer dependencies on multi-layered networks. The key idea behind our method is to view it as a collective collaborative filtering problem. By formulating the problem into a regularized optimization model, we propose an effective algorithm to find the local optima with linear complexity. Furthermore, we derive an online algorithm to accommodate newly arrived nodes, whose complexity is just linear wrt the size of the neighborhood of the new node. We perform extensive empirical evaluations to demonstrate the effectiveness and the efficiency of the proposed methods.

BACKGROUND: The recent outbreak of Ebola has been cited as the largest in history. Despite this global health crisis, few drugs are available to efficiently treat Ebola infections. Drug repurposing provides a potentially efficient solution to accelerating the development of therapeutic approaches in response to Ebola outbreak. To identify such candidates, we use an integrated structural systems pharmacology pipeline which combines proteome-scale ligand binding site comparison, protein-ligand docking, and Molecular Dynamics (MD) simulation. RESULTS: One thousand seven hundred and sixty-six FDA-approved drugs and 259 experimental drugs were screened to identify those with the potential to inhibit the replication and virulence of Ebola, and to determine the binding modes with their respective targets. Initial screening has identified a number of promising hits. Notably, Indinavir; an HIV protease inhibitor, may be effective in reducing the virulence of Ebola. Additionally, an antifungal (Sinefungin) and several anti-viral drugs (e.g. Maraviroc, Abacavir, Telbivudine, and Cidofovir) may inhibit Ebola RNA-directed RNA polymerase through targeting the MTase domain. CONCLUSIONS: Identification of safe drug candidates is a crucial first step toward the determination of timely and effective therapeutic approaches to address and mitigate the impact of the Ebola global crisis and future outbreaks of pathogenic diseases. Further in vitro and in vivo testing to evaluate the anti-Ebola activity of these drugs is warranted.

Target-based screening is one of the major approaches in drug discovery. Besides the intended target, unexpected drug off-target interactions often occur, and many of them have not been recognized and characterized. The off-target interactions can be responsible for either therapeutic or side effects. Thus, identifying the genome-wide off-targets of lead compounds or existing drugs will be critical for designing effective and safe drugs, and providing new opportunities for drug repurposing. Although many computational methods have been developed to predict drug-target interactions, they are either less accurate than the one that we are proposing here or computationally too intensive, thereby limiting their capability for large-scale off-target identification. In addition, the performances of most machine learning based algorithms have been mainly evaluated to predict off-target interactions in the same gene family for hundreds of chemicals. It is not clear how these algorithms perform in terms of detecting off-targets across gene families on a proteome scale. Here, we are presenting a fast and accurate off-target prediction method, REMAP, which is based on a dual regularized one-class collaborative filtering algorithm, to explore continuous chemical space, protein space, and their interactome on a large scale. When tested in a reliable, extensive, and cross-gene family benchmark, REMAP outperforms the state-of-the-art methods. Furthermore, REMAP is highly scalable. It can screen a dataset of 200 thousands chemicals against 20 thousands proteins within 2 hours. Using the reconstructed genome-wide target profile as the fingerprint of a chemical compound, we predicted that seven FDA-approved drugs can be repurposed as novel anti-cancer therapies. The anti-cancer activity of six of them is supported by experimental evidences. Thus, REMAP is a valuable addition to the existing in silico toolbox for drug target identification, drug repurposing, phenotypic screening, and side effect prediction. The software and benchmark are available at https://github.com/hansaimlim/REMAP.

MOTIVATION: Similarity search is the foundation of bioinformatics. It plays a key role in establishing structural, functional and evolutionary relationships between biological sequences. Although the power of the similarity search has increased steadily in recent years, a high percentage of sequences remain uncharacterized in the protein universe. Thus, new similarity search strategies are needed to efficiently and reliably infer the structure and function of new sequences. The existing paradigm for studying protein sequence, structure, function and evolution has been established based on the assumption that the protein universe is discrete and hierarchical. Cumulative evidence suggests that the protein universe is continuous. As a result, conventional sequence homology search methods may be not able to detect novel structural, functional and evolutionary relationships between proteins from weak and noisy sequence signals. To overcome the limitations in existing similarity search methods, we propose a new algorithmic framework-Enrichment of Network Topological Similarity (ENTS)-to improve the performance of large scale similarity searches in bioinformatics. RESULTS: We apply ENTS to a challenging unsolved problem: protein fold recognition. Our rigorous benchmark studies demonstrate that ENTS considerably outperforms state-of-the-art methods. As the concept of ENTS can be applied to any similarity metric, it may provide a general framework for similarity search on any set of biological entities, given their representation as a network. AVAILABILITY AND IMPLEMENTATION: Source code freely available upon request CONTACT: : [email protected].

Genome-Wide Association Studies (GWAS), whole genome sequencing, and high-throughput omics techniques have generated vast amounts of genotypic and molecular phenotypic data. However, these data have not yet been fully explored to improve the effectiveness and efficiency of drug discovery, which continues along a one-drug-one-target-one-disease paradigm. As a partial consequence, both the cost to launch a new drug and the attrition rate are increasing. Systems pharmacology and pharmacogenomics are emerging to exploit the available data and potentially reverse this trend, but, as we argue here, more is needed. To understand the impact of genetic, epigenetic, and environmental factors on drug action, we must study the structural energetics and dynamics of molecular interactions in the context of the whole human genome and interactome. Such an approach requires an integrative modeling framework for drug action that leverages advances in data-driven statistical modeling and mechanism-based multiscale modeling and transforms heterogeneous data from GWAS, high-throughput sequencing, structural genomics, functional genomics, and chemical genomics into unified knowledge. This is not a small task, but, as reviewed here, progress is being made towards the final goal of personalized medicines for the treatment of complex diseases.

Nelfinavir is a potent HIV-protease inhibitor with pleiotropic effects in cancer cells. Experimental studies connect its anti-cancer effects to the suppression of the Akt signaling pathway, but the actual molecular targets remain unknown. Using a structural proteome-wide off-target pipeline, which integrates molecular dynamics simulation and MM/GBSA free energy calculations with ligand binding site comparison and biological network analysis, we identified putative human off-targets of Nelfinavir and analyzed the impact on the associated biological processes. Our results suggest that Nelfinavir is able to inhibit multiple members of the protein kinase-like superfamily, which are involved in the regulation of cellular processes vital for carcinogenesis and metastasis. The computational predictions are supported by kinase activity assays and are consistent with existing experimental and clinical evidence. This finding provides a molecular basis to explain the broad-spectrum anti-cancer effect of Nelfinavir and presents opportunities to optimize the drug as a targeted polypharmacology agent.

Recent advances in structural bioinformatics have enabled the prediction of protein-drug off-targets based on their ligand binding sites. Concurrent developments in systems biology allow for prediction of the functional effects of system perturbations using large-scale network models. Integration of these two capabilities provides a framework for evaluating metabolic drug response phenotypes in silico. This combined approach was applied to investigate the hypertensive side effect of the cholesteryl ester transfer protein inhibitor torcetrapib in the context of human renal function. A metabolic kidney model was generated in which to simulate drug treatment. Causal drug off-targets were predicted that have previously been observed to impact renal function in gene-deficient patients and may play a role in the adverse side effects observed in clinical trials. Genetic risk factors for drug treatment were also predicted that correspond to both characterized and unknown renal metabolic disorders as well as cryptic genetic deficiencies that are not expected to exhibit a renal disorder phenotype except under drug treatment. This study represents a novel integration of structural and systems biology and a first step towards computational systems medicine. The methodology introduced herein has important implications for drug development and personalized medicine.

Conventional drug design embraces the "one gene, one drug, one disease" philosophy. Polypharmacology, which focuses on multi-target drugs, has emerged as a new paradigm in drug discovery. The rational design of drugs that act via polypharmacological mechanisms can produce compounds that exhibit increased therapeutic potency and against which resistance is less likely to develop. Additionally, identifying multiple protein targets is also critical for side-effect prediction. One third of potential therapeutic compounds fail in clinical trials or are later removed from the market due to unacceptable side effects often caused by off-target binding. In the current work, we introduce a multidimensional strategy for the identification of secondary targets of known small-molecule inhibitors in the absence of global structural and sequence homology with the primary target protein. To demonstrate the utility of the strategy, we identify several targets of 4,5-dihydroxy-3-(1-naphthyldiazenyl)-2,7-naphthalenedisulfonic acid, a known micromolar inhibitor of Trypanosoma brucei RNA editing ligase 1. As it is capable of identifying potential secondary targets, the strategy described here may play a useful role in future efforts to reduce drug side effects and/or to increase polypharmacology.

We report a computational approach that integrates structural bioinformatics, molecular modelling and systems biology to construct a drug-target network on a structural proteome-wide scale. The approach has been applied to the genome of Mycobacterium tuberculosis (M.tb), the causative agent of one of today's most widely spread infectious diseases. The resulting drug-target interaction network for all structurally characterized approved drugs bound to putative M.tb receptors, we refer to as the 'TB-drugome'. The TB-drugome reveals that approximately one-third of the drugs examined have the potential to be repositioned to treat tuberculosis and that many currently unexploited M.tb receptors may be chemically druggable and could serve as novel anti-tubercular targets. Furthermore, a detailed analysis of the TB-drugome has shed new light on the controversial issues surrounding drug-target networks [1]-[3]. Indeed, our results support the idea that drug-target networks are inherently modular, and further that any observed randomness is mainly caused by biased target coverage. The TB-drugome (http://funsite.sdsc.edu/drugome/TB) has the potential to be a valuable resource in the development of safe and efficient anti-tubercular drugs. More generally the methodology may be applied to other pathogens of interest with results improving as more of their structural proteomes are determined through the continued efforts of structural biology/genomics.

Functional relationships between proteins that do not share global structure similarity can be established by detecting their ligand-binding-site similarity. For a large-scale comparison, it is critical to accurately and efficiently assess the statistical significance of this similarity. Here, we report an efficient statistical model that supports local sequence order independent ligand-binding-site similarity searching. Most existing statistical models only take into account the matching vertices between two sites that are defined by a fixed number of points. In reality, the boundary of the binding site is not known or is dependent on the bound ligand making these approaches limited. To address these shortcomings and to perform binding-site mapping on a genome-wide scale, we developed a sequence-order independent profile-profile alignment (SOIPPA) algorithm that is able to detect local similarity between unknown binding sites a priori. The SOIPPA scoring integrates geometric, evolutionary and physical information into a unified framework. However, this imposes a significant challenge in assessing the statistical significance of the similarity because the conventional probability model that is based on fixed-point matching cannot be applied. Here we find that scores for binding-site matching by SOIPPA follow an extreme value distribution (EVD). Benchmark studies show that the EVD model performs at least two-orders faster and is more accurate than the non-parametric statistical method in the previous SOIPPA version. Efficient statistical analysis makes it possible to apply SOIPPA to genome-based drug discovery. Consequently, we have applied the approach to the structural genome of Mycobacterium tuberculosis to construct a protein-ligand interaction network. The network reveals highly connected proteins, which represent suitable targets for promiscuous drugs.

The rise of multi-drug resistant (MDR) and extensively drug resistant (XDR) tuberculosis around the world, including in industrialized nations, poses a great threat to human health and defines a need to develop new, effective and inexpensive anti-tubercular agents. Previously we developed a chemical systems biology approach to identify off-targets of major pharmaceuticals on a proteome-wide scale. In this paper we further demonstrate the value of this approach through the discovery that existing commercially available drugs, prescribed for the treatment of Parkinson's disease, have the potential to treat MDR and XDR tuberculosis. These drugs, entacapone and tolcapone, are predicted to bind to the enzyme InhA and directly inhibit substrate binding. The prediction is validated by in vitro and InhA kinetic assays using tablets of Comtan, whose active component is entacapone. The minimal inhibition concentration (MIC(99)) of entacapone for Mycobacterium tuberculosis (M.tuberculosis) is approximately 260.0 microM, well below the toxicity concentration determined by an in vitro cytotoxicity model using a human neuroblastoma cell line. Moreover, kinetic assays indicate that Comtan inhibits InhA activity by 47.0% at an entacapone concentration of approximately 80 microM. Thus the active component in Comtan represents a promising lead compound for developing a new class of anti-tubercular therapeutics with excellent safety profiles. More generally, the protocol described in this paper can be included in a drug discovery pipeline in an effort to discover novel drug leads with desired safety profiles, and therefore accelerate the development of new drugs.

Systematic identification of protein-drug interaction networks is crucial to correlate complex modes of drug action to clinical indications. We introduce a novel computational strategy to identify protein-ligand binding profiles on a genome-wide scale and apply it to elucidating the molecular mechanisms associated with the adverse drug effects of Cholesteryl Ester Transfer Protein (CETP) inhibitors. CETP inhibitors are a new class of preventive therapies for the treatment of cardiovascular disease. However, clinical studies indicated that one CETP inhibitor, Torcetrapib, has deadly off-target effects as a result of hypertension, and hence it has been withdrawn from phase III clinical trials. We have identified a panel of off-targets for Torcetrapib and other CETP inhibitors from the human structural genome and map those targets to biological pathways via the literature. The predicted protein-ligand network is consistent with experimental results from multiple sources and reveals that the side-effect of CETP inhibitors is modulated through the combinatorial control of multiple interconnected pathways. Given that combinatorial control is a common phenomenon observed in many biological processes, our findings suggest that adverse drug effects might be minimized by fine-tuning multiple off-target interactions using single or multiple therapies. This work extends the scope of chemogenomics approaches and exemplifies the role that systems biology has in the future of drug discovery.

BACKGROUND: An accurate description of protein shape derived from protein structure is necessary to establish an understanding of protein-ligand interactions, which in turn will lead to improved methods for protein-ligand docking and binding site analysis. Most current shape descriptors characterize only the local properties of protein structure using an all-atom representation and are slow to compute. We need new shape descriptors that have the ability to capture both local and global structural information, are robust for application to models and low quality structures and are computationally efficient to permit high throughput analysis of protein structures. RESULTS: We introduce a new shape description that requires only the Calpha atoms to represent the protein structure, thus making it both fast and suitable for use on models and low quality structures. The notion of a geometric potential is introduced to quantitatively describe the shape of the structure. This geometric potential is dependent on both the global shape of the protein structure as well as the surrounding environment of each residue. When applying the geometric potential for binding site prediction, approximately 85% of known binding sites can be accurately identified with above 50% residue coverage and 80% specificity. Moreover, the algorithm is fast enough for proteome-scale applications. Proteins with fewer than 500 amino acids can be scanned in less than two seconds. CONCLUSION: The reduced representation of the protein structure combined with the geometric potential provides a fast, quantitative description of protein-ligand binding sites with potential for use in large-scale predictions, comparisons and analysis.

Early identification of adverse effect of preclinical and commercial drugs is crucial in developing highly efficient therapeutics, since unexpected adverse drug effects account for one-third of all drug failures in drug development. To correlate protein-drug interactions at the molecule level with their clinical outcomes at the organism level, we have developed an integrated approach to studying protein-ligand interactions on a structural proteome-wide scale by combining protein functional site similarity search, small molecule screening, and protein-ligand binding affinity profile analysis. By applying this methodology, we have elucidated a possible molecular mechanism for the previously observed, but molecularly uncharacterized, side effect of selective estrogen receptor modulators (SERMs). The side effect involves the inhibition of the Sacroplasmic Reticulum Ca2+ ion channel ATPase protein (SERCA) transmembrane domain. The prediction provides molecular insight into reducing the adverse effect of SERMs and is supported by clinical and in vitro observations. The strategy used in this case study is being applied to discover off-targets for other commercially available pharmaceuticals. The process can be included in a drug discovery pipeline in an effort to optimize drug leads and reduce unwanted side effects.

The bias in protein structure and function space resulting from experimental limitations and targeting of particular functional classes of proteins by structural biologists has long been recognized, but never continuously quantified. Using the Enzyme Commission and the Gene Ontology classifications as a reference frame, and integrating structure data from the Protein Data Bank (PDB), target sequences from the structural genomics projects, structure homology derived from the SUPERFAMILY database, and genome annotations from Ensembl and NCBI, we provide a quantified view, both at the domain and whole-protein levels, of the current and projected coverage of protein structure and function space relative to the human genome. Protein structures currently provide at least one domain that covers 37% of the functional classes identified in the genome; whole structure coverage exists for 25% of the genome. If all the structural genomics targets were solved (twice the current number of structures in the PDB), it is estimated that structures of one domain would cover 69% of the functional classes identified and complete structure coverage would be 44%. Homology models from existing experimental structures extend the 37% coverage to 56% of the genome as single domains and 25% to 31% for complete structures. Coverage from homology models is not evenly distributed by protein family, reflecting differing degrees of sequence and structure divergence within families. While these data provide coverage, conversely, they also systematically highlight functional classes of proteins for which structures should be determined. Current key functional families without structure representation are highlighted here; updated information on the "most wanted list" that should be solved is available on a weekly basis from http://function.rcsb.org:8080/pdb/function_distribution/index.html.