Drug-target interaction prediction by random walk on the heterogeneous network

Predicting potential drug-target interactions from heterogeneous biological data is critical not only for better understanding of the various interactions and biological processes, but also for the development of novel drugs and the improvement of human medicines. In this paper, the method of Network-based Random Walk with Restart on the Heterogeneous network (NRWRH) is developed to predict potential drug-target interactions on a large scale under the hypothesis that similar drugs often target similar target proteins and the framework of Random Walk. Compared with traditional supervised or semi-supervised methods, NRWRH makes full use of the tool of the network for data integration to predict drug-target associations. It integrates three different networks (protein-protein similarity network, drug-drug similarity network, and known drug-target interaction networks) into a heterogeneous network by known drug-target interactions and implements the random walk on this heterogeneous network. When applied to four classes of important drug-target interactions including enzymes, ion channels, GPCRs and nuclear receptors, NRWRH significantly improves previous methods in terms of cross-validation and potential drug-target interaction prediction. Excellent performance enables us to suggest a number of new potential drug-target interactions for drug development.     

Drug-target interaction prediction via class imbalance-aware ensemble learning

Background

Multiple computational methods for predicting drug-target interactions have been developed to facilitate the drug discovery process. These methods use available data on known drug-target interactions to train classifiers with the purpose of predicting new undiscovered interactions. However, a key challenge regarding this data that has not yet been addressed by these methods, namely class imbalance, is potentially degrading the prediction performance. Class imbalance can be divided into two sub-problems. Firstly, the number of known interacting drug-target pairs is much smaller than that of non-interacting drug-target pairs. This imbalance ratio between interacting and non-interacting drug-target pairs is referred to as the between-class imbalance. Between-class imbalance degrades prediction performance due to the bias in prediction results towards the majority class (i.e. the non-interacting pairs), leading to more prediction errors in the minority class (i.e. the interacting pairs). Secondly, there are multiple types of drug-target interactions in the data with some types having relatively fewer members (or are less represented) than others. This variation in representation of the different interaction types leads to another kind of imbalance referred to as the within-class imbalance. In within-class imbalance, prediction results are biased towards the better represented interaction types, leading to more prediction errors in the less represented interaction types.

Results

We propose an ensemble learning method that incorporates techniques to address the issues of between-class imbalance and within-class imbalance. Experiments show that the proposed method improves results over 4 state-of-the-art methods. In addition, we simulated cases for new drugs and targets to see how our method would perform in predicting their interactions. New drugs and targets are those for which no prior interactions are known. Our method displayed satisfactory prediction performance and was able to predict many of the interactions successfully.

Conclusions

Our proposed method has improved the prediction performance over the existing work, thus proving the importance of addressing problems pertaining to class imbalance in the data.

     

Enhancing protein-vitamin binding residues prediction by multiple heterogeneous subspace SVMs ensemble

Background

Vitamins are typical ligands that play critical roles in various metabolic processes. The accurate identification of the vitamin-binding residues solely based on a protein sequence is of significant importance for the functional annotation of proteins, especially in the post-genomic era, when large volumes of protein sequences are accumulating quickly without being functionally annotated.

Results

In this paper, a new predictor called TargetVita is designed and implemented for predicting protein-vitamin binding residues using protein sequences. In TargetVita, features derived from the position-specific scoring matrix (PSSM), predicted protein secondary structure, and vitamin binding propensity are combined to form the original feature space; then, several feature subspaces are selected by performing different feature selection methods. Finally, based on the selected feature subspaces, heterogeneous SVMs are trained and then ensembled for performing prediction.

Conclusions

The experimental results obtained with four separate vitamin-binding benchmark datasets demonstrate that the proposed TargetVita is superior to the state-of-the-art vitamin-specific predictor, and an average improvement of 10% in terms of the Matthews correlation coefficient (MCC) was achieved over independent validation tests. The TargetVita web server and the datasets used are freely available for academic use at http://csbio.njust.edu.cn/bioinf/TargetVita or http://www.csbio.sjtu.edu.cn/bioinf/TargetVita.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2105-15-297) contains supplementary material, which is available to authorized users.     

Open source tool for prediction of genome wide protein-protein interaction network based on ortholog information

Background  

Protein-protein interactions are crucially important for cellular processes. Knowledge of these interactions improves the understanding of cell cycle, metabolism, signaling, transport, and secretion. Information about interactions can hint at molecular causes of diseases, and can provide clues for new therapeutic approaches. Several (usually expensive and time consuming) experimental methods can probe protein - protein interactions. Data sets, derived from such experiments make the development of prediction methods feasible, and make the creation of protein-protein interaction network predicting tools possible.     

Semi-supervised drug-protein interaction prediction from heterogeneous biological spaces

Background

Predicting drug-protein interactions from heterogeneous biological data sources is a key step for in silico drug discovery. The difficulty of this prediction task lies in the rarity of known drug-protein interactions and myriad unknown interactions to be predicted. To meet this challenge, a manifold regularization semi-supervised learning method is presented to tackle this issue by using labeled and unlabeled information which often generates better results than using the labeled data alone. Furthermore, our semi-supervised learning method integrates known drug-protein interaction network information as well as chemical structure and genomic sequence data.

Results

Using the proposed method, we predicted certain drug-protein interactions on the enzyme, ion channel, GPCRs, and nuclear receptor data sets. Some of them are confirmed by the latest publicly available drug targets databases such as KEGG.

Conclusions

We report encouraging results of using our method for drug-protein interaction network reconstruction which may shed light on the molecular interaction inference and new uses of marketed drugs.

     

Protein complexes identification based on go attributed network embedding

Background

Identifying protein complexes from protein-protein interaction (PPI) network is one of the most important tasks in proteomics. Existing computational methods try to incorporate a variety of biological evidences to enhance the quality of predicted complexes. However, it is still a challenge to integrate different types of biological information into the complexes discovery process under a unified framework. Recently, attributed network embedding methods have be proved to be remarkably effective in generating vector representations for nodes in the network. In the transformed vector space, both the topological proximity and node attributed affinity between different nodes are preserved. Therefore, such attributed network embedding methods provide us a unified framework to integrate various biological evidences into the protein complexes identification process.

Results

In this article, we propose a new method called GANE to predict protein complexes based on Gene Ontology (GO) attributed network embedding. Firstly, it learns the vector representation for each protein from a GO attributed PPI network. Based on the pair-wise vector representation similarity, a weighted adjacency matrix is constructed. Secondly, it uses the clique mining method to generate candidate cores. Consequently, seed cores are obtained by ranking candidate cores based on their densities on the weighted adjacency matrix and removing redundant cores. For each seed core, its attachments are the proteins with correlation score that is larger than a given threshold. The combination of a seed core and its attachment proteins is reported as a predicted protein complex by the GANE algorithm. For performance evaluation, we compared GANE with six protein complex identification methods on five yeast PPI networks. Experimental results showes that GANE performs better than the competing algorithms in terms of different evaluation metrics.

Conclusions

GANE provides a framework that integrate many valuable and different biological information into the task of protein complex identification. The protein vector representation learned from our attributed PPI network can also be used in other tasks, such as PPI prediction and disease gene prediction.

     

Procedures for selecting composites based on prediction methods

Summary Procedures for selecting among parental varieties to be used in the synthesis of composites are discussed. In addition to the criterion based on the mean and variance of composites of the same size (k) proposed by Cordoso (1976), we suggest the index I_j=w₁v_j+w_{2 j} or I_j=(2/k) I_j for a preliminary selection among parental varieties. We show that by increasing k (size of the composite) I_j tends to g_j, the general combining ability effect. Such a criterion is particularly important when n, the number of parental varieties, is large, so that the number of possible composites (N_c=2ⁿ–n–1) becomes too large to be handled when using the common prediction procedures. Yield data from a 9 × 9 variety diallel cross were used for illustration.     

Conserved network motifs allow protein-protein interaction prediction

MOTIVATION: High-throughput protein interaction detection methods are strongly affected by false positive and false negative results. Focused experiments are needed to complement the large-scale methods by validating previously detected interactions but it is often difficult to decide which proteins to probe as interaction partners. Developing reliable computational methods assisting this decision process is a pressing need in bioinformatics. RESULTS: We show that we can use the conserved properties of the protein network to identify and validate interaction candidates. We apply a number of machine learning algorithms to the protein connectivity information and achieve a surprisingly good overall performance in predicting interacting proteins. Using a 'leave-one-out' approach we find average success rates between 20 and 40% for predicting the correct interaction partner of a protein. We demonstrate that the success of these methods is based on the presence of conserved interaction motifs within the network. AVAILABILITY: A reference implementation and a table with candidate interacting partners for each yeast protein are available at http://www.protsuggest.org.     

Artificial intelligence based methods for hot spot prediction

Proteins interact through their interfaces to fulfill essential functions in the cell. They bind to their partners in a highly specific manner and form complexes that have a profound effect on understanding the biological pathways they are involved in. Any abnormal interactions may cause diseases. Therefore, the identification of small molecules which modulate protein interactions through their interfaces has high therapeutic potential. However, discovering such molecules is challenging. Most protein–protein binding affinity is attributed to a small set of amino acids found in protein interfaces known as hot spots. Recent studies demonstrate that drug-like small molecules specifically may bind to hot spots. Therefore, hot spot prediction is crucial. As experimental data accumulates, artificial intelligence begins to be used for computational hot spot prediction. First, we review machine learning and deep learning for computational hot spot prediction and then explain the significance of hot spots toward drug design.     

Gene gravity-like algorithm for disease gene prediction based on phenotype-specific network

Background

Polygenic diseases are usually caused by the dysfunction of multiple genes. Unravelling such disease genes is crucial to fully understand the genetic landscape of diseases on molecular level. With the advent of ‘omic’ data era, network-based methods have prominently boosted disease gene discovery. However, how to make better use of different types of data for the prediction of disease genes remains a challenge.

Results

In this study, we improved the performance of disease gene prediction by integrating the similarity of disease phenotype, biological function and network topology. First, for each phenotype, a phenotype-specific network was specially constructed by mapping phenotype similarity information of given phenotype onto the protein-protein interaction (PPI) network. Then, we developed a gene gravity-like algorithm, to score candidate genes based on not only topological similarity but also functional similarity. We tested the proposed network and algorithm by conducting leave-one-out and leave-10%-out cross validation and compared them with state-of-art algorithms. The results showed a preference to phenotype-specific network as well as gene gravity-like algorithm. At last, we tested the predicting capacity of proposed algorithms by test gene set derived from the DisGeNET database. Also, potential disease genes of three polygenic diseases, obesity, prostate cancer and lung cancer, were predicted by proposed methods. We found that the predicted disease genes are highly consistent with literature and database evidence.

Conclusions

The good performance of phenotype-specific networks indicates that phenotype similarity information has positive effect on the prediction of disease genes. The proposed gene gravity-like algorithm outperforms the algorithm of Random Walk with Restart (RWR), implicating its predicting capacity by combing topological similarity with functional similarity. Our work will give an insight to the discovery of disease genes by fusing multiple similarities of genes and diseases.

     

Effect of reference genome selection on the performance of computational methods for genome-wide protein-protein interaction prediction

Background

Recent progress in computational methods for predicting physical and functional protein-protein interactions has provided new insights into the complexity of biological processes. Most of these methods assume that functionally interacting proteins are likely to have a shared evolutionary history. This history can be traced out for the protein pairs of a query genome by correlating different evolutionary aspects of their homologs in multiple genomes known as the reference genomes. These methods include phylogenetic profiling, gene neighborhood and co-occurrence of the orthologous protein coding genes in the same cluster or operon. These are collectively known as genomic context methods. On the other hand a method called mirrortree is based on the similarity of phylogenetic trees between two interacting proteins. Comprehensive performance analyses of these methods have been frequently reported in literature. However, very few studies provide insight into the effect of reference genome selection on detection of meaningful protein interactions.

Methods

We analyzed the performance of four methods and their variants to understand the effect of reference genome selection on prediction efficacy. We used six sets of reference genomes, sampled in accordance with phylogenetic diversity and relationship between organisms from 565 bacteria. We used Escherichia coli as a model organism and the gold standard datasets of interacting proteins reported in DIP, EcoCyc and KEGG databases to compare the performance of the prediction methods.

Conclusions

Higher performance for predicting protein-protein interactions was achievable even with 100–150 bacterial genomes out of 565 genomes. Inclusion of archaeal genomes in the reference genome set improves performance. We find that in order to obtain a good performance, it is better to sample few genomes of related genera of prokaryotes from the large number of available genomes. Moreover, such a sampling allows for selecting 50–100 genomes for comparable accuracy of predictions when computational resources are limited.     

Domain-mediated protein interaction prediction: From genome to network

Protein-protein interactions (PPIs), involved in many biological processes such as cellular signaling, are ultimately encoded in the genome. Solving the problem of predicting protein interactions from the genome sequence will lead to increased understanding of complex networks, evolution and human disease. We can learn the relationship between genomes and networks by focusing on an easily approachable subset of high-resolution protein interactions that are mediated by peptide recognition modules (PRMs) such as PDZ, WW and SH3 domains. This review focuses on computational prediction and analysis of PRM-mediated networks and discusses sequence- and structure-based interaction predictors, techniques and datasets for identifying physiologically relevant PPIs, and interpreting high-resolution interaction networks in the context of evolution and human disease.     

Novel link prediction for large-scale miRNA-lncRNA interaction network in a bipartite graph

Background

Current knowledge and data on miRNA-lncRNA interactions is still limited and little effort has been made to predict target lncRNAs of miRNAs. Accumulating evidences suggest that the interaction patterns between lncRNAs and miRNAs are closely related to relative expression level, forming a titration mechanism. It could provide an effective approach for characteristic feature extraction. In addition, using the coding non-coding co-expression network and sequence data could also help to measure the similarities among miRNAs and lncRNAs. By mathematically analyzing these types of similarities, we come up with two findings that (i) lncRNAs/miRNAs tend to collaboratively interact with miRNAs/lncRNAs of similar expression profiles, and vice versa, and (ii) those miRNAs interacting with a cluster of common target genes tend to jointly target at the common lncRNAs.

Methods

In this work, we developed a novel group preference Bayesian collaborative filtering model called GBCF for picking up a top-k probability ranking list for an individual miRNA or lncRNA based on the known miRNA-lncRNA interaction network.

Results

To evaluate the effectiveness of GBCF, leave-one-out and k-fold cross validations as well as a series of comparison experiments were carried out. GBCF achieved the values of area under ROC curve of 0.9193, 0.8354+/??0.0079, 0.8615+/??0.0078, and 0.8928+/??0.0082 based on leave-one-out, 2-fold, 5-fold, and 10-fold cross validations respectively, demonstrating its reliability and robustness.

Conclusions

GBCF could be used to select potential lncRNA targets of specific miRNAs and offer great insights for further researches on ceRNA regulation network.

     

Protein-protein interaction site prediction based on conditional random fields

MOTIVATION: We are motivated by the fast-growing number of protein structures in the Protein Data Bank with necessary information for prediction of protein-protein interaction sites to develop methods for identification of residues participating in protein-protein interactions. We would like to compare conditional random fields (CRFs)-based method with conventional classification-based methods that omit the relation between two labels of neighboring residues to show the advantages of CRFs-based method in predicting protein-protein interaction sites. RESULTS: The prediction of protein-protein interaction sites is solved as a sequential labeling problem by applying CRFs with features including protein sequence profile and residue accessible surface area. The CRFs-based method can achieve a comparable performance with state-of-the-art methods, when 1276 nonredundant hetero-complex protein chains are used as training and test set. Experimental result shows that CRFs-based method is a powerful and robust protein-protein interaction site prediction method and can be used to guide biologists to make specific experiments on proteins. AVAILABILITY: http://www.insun.hit.edu.cn/~mhli/site_CRFs/index.html. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.     

Quantitative prediction of miRNA-mRNA interaction based on equilibrium concentrations

MicroRNAs (miRNAs) suppress gene expression by forming a duplex with a target messenger RNA (mRNA), blocking translation or initiating cleavage. Computational approaches have proven valuable for predicting which mRNAs can be targeted by a given miRNA, but currently available prediction methods do not address the extent of duplex formation under physiological conditions. Some miRNAs can at low concentrations bind to target mRNAs, whereas others are unlikely to bind within a physiologically relevant concentration range. Here we present a novel approach in which we find potential target sites on mRNA that minimize the calculated free energy of duplex formation, compute the free energy change involved in unfolding these sites, and use these energies to estimate the extent of duplex formation at specified initial concentrations of both species. We compare our predictions to experimentally confirmed miRNA-mRNA interactions (and non-interactions) in Drosophila melanogaster and in human. Although our method does not predict whether the targeted mRNA is degraded and/or its translation to protein inhibited, our quantitative estimates generally track experimentally supported results, indicating that this approach can be used to predict whether an interaction occurs at specified concentrations. Our approach offers a more-quantitative understanding of post-translational regulation in different cell types, tissues, and developmental conditions.     

Predicting protein phenotypes based on protein-protein interaction network

Background

Identifying associated phenotypes of proteins is a challenge of the modern genetics since the multifactorial trait often results from contributions of many proteins. Besides the high-through phenotype assays, the computational methods are alternative ways to identify the phenotypes of proteins.

Methodology/Principal Findings

Here, we proposed a new method for predicting protein phenotypes in yeast based on protein-protein interaction network. Instead of only the most likely phenotype, a series of possible phenotypes for the query protein were generated and ranked acording to the tethering potential score. As a result, the first order prediction accuracy of our method achieved 65.4% evaluated by Jackknife test of 1,267 proteins in budding yeast, much higher than the success rate (15.4%) of a random guess. And the likelihood of the first 3 predicted phenotypes including all the real phenotypes of the proteins was 70.6%.

Conclusions/Significance

The candidate phenotypes predicted by our method provided useful clues for the further validation. In addition, the method can be easily applied to the prediction of protein associated phenotypes in other organisms.