Database Of Interacting Proteins Research Articles

Decision tree classifier based on topological characteristics of subgraph for the mining of protein complexes from large scale PPI networks

The growing accessibility of large-scale protein interaction data demands extensive research to understand cell organization and its functioning at the network level. Bioinformatics and data mining researchers have extensively studied network clustering to examine the structural and operational features of protein protein interaction (PPI) networks. Clustering PPI networks has proven useful in numerous research over the past two decades for identifying functional modules, understanding the roles of previously unknown proteins, and other purposes. Protein complexes represent one of the essential cellular components for creating biological activities. Inferring protein complexes has been made more accessible by experimental approaches. We offer a novel method that integrates the classification model with local topological data, making it more reliable and efficient. This article describes a decision tree classifier based on topological characteristics of the subgraph for mining protein complexes. The proposed graph-based algorithm is an effective and efficient way to identify protein complexes from large-scale PPI networks. The performance of the proposed algorithm is observed in protein–protein interaction networks of yeast and human in the Database of Interacting Proteins (DIP) and the Biological General Repository for Interaction Datasets (BioGRID) using widely accepted benchmark protein complexes from the comprehensive resource of mammalian protein complexes (CORUM) and the comprehensive catalogue of yeast protein complexes (CYC2008). The outcomes demonstrate that our method can outperform the best-performing supervised, semi-supervised, and unsupervised approaches to detecting protein complexes.

Additional Pathogenic Pathways in RBCK1 Deficiency

RBCK1 deficiency is a rare congenital autoinflammatory disease that causes inflammatory disruption on the molecular level. This deficiency has three major clinical manifestations: increased sensitivity to bacterial infections, autoinflammation syndrome, and the accumulation of amylopectin in skeletal muscle. The amylopectinosis causes myopathy and cardiomyopathy. The pathogenesis of the disease is poorly investigated and may include unnoticed relationships. We performed gene expression analysis on patients with RBCK1 deficiency and three other autoinflammatory diseases. The identification of differentially expressed genes revealed a large number of downregulated genes that are involved in the activation of essential metabolic and immune pathways, including NF-kB and Pi3k-Akt-mTOR. Signaling pathways were analysed using the KEGG (Kyoto Encyclopedia of Genes and Genomes) and Gene Ontology resource. Predicted protein-protein interactions were retrieved from the STRING (Search Tool for the Retrieval of Interacting proteins database). Besides the primary involvement of RBCK1 in disease pathology, several downregulated pathways aggravate symptoms of myopathy, cardiomyopathy, and bacterial disease. The studied pathways may serve as new targets for the development of compensatory therapies for patients with RBCK1 deficiency.

Establishment of an Endocytosis-Related Prognostic Signature for Patients With Low-Grade Glioma.

BackgroundLow-grade glioma (LGG) is a heterogeneous tumor that might develop into high-grade malignant glioma, which markedly reduces patient survival time. Endocytosis is a cellular process responsible for the internalization of cell surface proteins or external materials into the cytosol. Dysregulated endocytic pathways have been linked to all steps of oncogenesis, from initial transformation to late invasion and metastasis. However, endocytosis-related gene (ERG) signatures have not been used to study the correlations between endocytosis and prognosis in cancer. Therefore, it is essential to develop a prognostic model for LGG based on the expression profiles of ERGs.MethodsThe Cancer Genome Atlas and the Genotype-Tissue Expression database were used to identify differentially expressed ERGs in LGG patients. Gene ontology, Kyoto Encyclopedia of Genes and Genomes, and Gene set enrichment analysis methodologies were adopted for functional analysis. A protein-protein interaction (PPI) network was constructed and hub genes were identified based on the Search Tool for the Retrieval of Interacting Proteins database. Univariate and multivariate Cox regression analyses were used to develop an ERG signature to predict the overall survival (OS) of LGG patients. Finally, the association between the ERG signature and gene mutation status was further analyzed.ResultsSixty-two ERGs showed distinct mRNA expression patterns between normal brain tissues and LGG tissues. Functional analysis indicated that these ERGs were strikingly enriched in endosomal trafficking pathways. The PPI network indicated that EGFR was the most central protein. We then built a 29-gene signature, dividing patients into high-risk and low-risk groups with significantly different OS times. The prognostic performance of the 29-gene signature was validated in another LGG cohort. Additionally, we found that the mutation scores calculated based on the TTN, PIK3CA, NF1, and IDH1 mutation status were significantly correlated with the endocytosis-related prognostic signature. Finally, a clinical nomogram with a concordance index of 0.881 predicted the survival probability of LGG patients by integrating clinicopathologic features and ERG signatures.ConclusionOur ERG-based prediction models could serve as an independent prognostic tool to accurately predict the outcomes of LGG.

A New Sequential Forward Feature Selection (SFFS) Algorithm for Mining Best Topological and Biological Features to Predict Protein Complexes from Protein–Protein Interaction Networks (PPINs)

Protein-protein interaction plays an important role in the understanding of biological processes in the body. A network of dynamic protein complexes within a cell that regulates most biological processes is known as a protein-protein interaction network (PPIN). Complex prediction from PPINs is a challenging task. Most of the previous computation approaches mine cliques, stars, linear and hybrid structures as complexes from PPINs by considering topological features and fewer of them focus on important biological information contained within protein amino acid sequence. In this study, we have computed a wide variety of topological features and integrate them with biological features computed from protein amino acid sequence such as bag of words, physicochemical and spectral domain features. We propose a new Sequential Forward Feature Selection (SFFS) algorithm, i.e., random forest-based Boruta feature selection for selecting the best features from computed large feature set. Decision tree, linear discriminant analysis and gradient boosting classifiers are used as learners. We have conducted experiments by considering two reference protein complex datasets of yeast, i.e., CYC2008 and MIPS. Human and mouse complex information is taken from CORUM 3.0 dataset. Protein interaction information is extracted from the database of interacting proteins (DIP). Our proposed SFFS, i.e., random forest-based Brouta feature selection in combination with decision trees, linear discriminant analysis and Gradient Boosting Classifiers outperforms other state of art algorithms by achieving precision, recall and F-measure rates, i.e. 94.58%, 94.92% and 94.45% for MIPS, 96.31%, 93.55% and 96.02% for CYC2008, 98.84%, 98.00%, 98.87 % for CORUM humans and 96.60%, 96.70%, 96.32% for CORUM mouse dataset complexes, respectively.

Network Pharmacology-Based Analysis of the Underlying Mechanism of Huajiao for Pain Relief.

Objective Pain is a common symptom among patients, and pain management is an important clinical practice topic. The mechanism of Huajiao (HJ; Zanthoxylum bungeanum Maxim.) and its effective components for treating pain was explored using network pharmacology and molecular docking to verify its pain relief function in traditional medical practice. Methods HJ's components were collected via the Traditional Chinese Medicine Systems Pharmacology platform and published studies. HJ-associated target proteins were predicted using the drug similarity rule via Swiss Target Prediction. Online Mendelian Inheritance in Man was used to search for pain-related genes and proteins, and the Database of Interacting Proteins was used to obtain the human interactive target proteins. The compound-target-disease network of HJ for pain relief was constructed with protein-protein interaction networks. The obtained target proteins were uploaded on the Database for Annotation, Visualization, and Integrated Discovery to annotate, visualize, and integrally discover the related signaling pathway, and semiflexible molecular docking by Autodock Vina was applied to verify the potential mechanism. Results A total of 157 molecules in HJ were obtained, and the top 20 active components or active groups were mainly focused on the amide alkaloids (e.g., [6RS]-[2E,7E,9E]-6-hydroxy-N-[2-hydroxy-2-methylpropyl]-11-oxo-2,7,9-dodecatrienamide and [2E,7E,9E]-N-[2-hydroxy-2-methylpropyl]-11-ethoxy-6-hydroxy-dodeca-2,7,9-trienamide). Also, the 66 main targets were filtered from 746 predicted targets and 928 pain-related targets through module Network Analyzer in Cytoscape 3.6.0. Finally, there were 3 critical signaling pathways, including mitogen-activated protein kinase, phosphoinositide 3-kinase-protein kinase B-mammalian target of rapamycin, and IκB kinase-nuclear factor κB-cyclooxygenase 2 based on integrated discovery with 54 enriched signaling pathways. Conclusions HJ is used as a pain relief and has multicomponents, multitargets, and multiapproaches. Amide alkaloids are important substance bases, and HJ is more suitable for treating inflammatory pain.

Algorithm for predicting weighted protein complexes by using modularity function

Due to the difficulties in obtaining the identification results of protein complexes with higher accuracy, as well as the running efficiency of the protein complexes identification algorithm based on weighted modularity function clustering is low, a novel protein complex recognition algorithm named IWPC-MF (Algorithm for identifying weighted protein complexes based on modularity function) was proposed. Firstly, we build the TCEWS (topological clustering for weighted edges) algorithm, which is founded on improved edge clustering coefficient and the combination of Pearson correlation coefficient with edge point clustering coefficient. Secondly, seed nodes were selected according to the weight of nodes, then the IMS (initial module selection strategy) program has been proposed. The similarity measurement and the protein attachment degree between nodes were selected to obtain the initial clustering module by traversing neighbors of seeds. Finally, based on the tightness, we take the IMC (initial module combination strategy) module with modularity function was added to initial module and finally complete the protein complex detection, which can overcome the defect that the traditional modularity function cannot identify the overlapped and small complex. IWPC-MF algorithm was used to identify protein complexes on DIP (the Database of Interacting Proteins) data, in order to reduce effectively the computational complexity and randomness, speed up the clustering speed. The experimental results show that IWPC-MF algorithm has better performance on accuracy and recall, which is more reasonable to identify protein complexes.

Establishment of a SVM classifier to predict recurrence of ovarian cancer.

Gene expression data using retrieved ovarian cancer (OC) samples were used to identify genes of interest and a support vector machine (SVM) classifier was subsequently established to predict the recurrence of OC. Three datasets (GSE17260, GSE44104 and GSE51088) investigating OC gene expression were downloaded from the Gene Expression Omnibus. Differentially expressed genes (DEGs) in samples from patients with non-recurrent and recurrent OC were revealed via a homogeneity test and quality control analysis. A protein-protein interaction (PPI) network was subsequently established for the DEGs using data from Biological General Repository for Interaction Datasets, Human Protein Reference Database and Database of Interacting Proteins. Degrees of interaction and betweenness centrality (BC) scores were calculated for each node in the PPI network. The top 100 genes ranked by BC scores were selected to identify feature genes via recursive feature elimination using the GSE17260 dataset. Following this, a SVM classifier was constructed and further validated using the GSE44104 and GSE51088 datasets and independent gene expression data obtained from the Cancer Genome Atlas (TCGA). A total of 639 DEGs were identified from the three gene expression datasets, and a PPI network including 249 nodes and 354 edges was constructed. A SVM classifier consisting of 39 feature genes (including cullin 3, mouse double minute 2 homolog, aurora kinase A, WW domain containing oxidoreducatase, large tumor suppressor kinase 2, sirtuin 6, staphylococcal nuclease and tudor domain containing 1, leucine rich repeats and immunoglobulin like domains 1 and aurora kinase 1 interacting protein 1) was subsequently constructed. The prediction accuracies of the SVM classifier for GSE17260, GSE44104 and GSE51088 datasets as well as data downloaded from TCGA were revealed to be 92.7, 93.3, 96.6 and 90.4%, respectively. Furthermore, the results of the present study revealed that patients with predicted non-recurrent OC survived significantly longer compared with the patients with predicted recurrent OC (P=6.598×10−6). A SVM classifier consisting of 39 feature genes was established for predicting the recurrence and prognosis of OC. Therefore, the results of the present study suggested that the 39 feature genes may serve important roles in the development of OC and may represent therapeutic biomarkers of OC.

Identification of Histone Deacetylase (HDAC) as a drug target against MRSA via interolog method of protein-protein interaction prediction

Patently, Protein-Protein Interactions (PPIs) lie at the core of significant biological functions and make the foundation of host-pathogen relationships. Hence, the current study is aimed to use computational biology techniques to predict host-pathogen Protein-Protein Interactions (HP-PPIs) between MRSA and Humans as potential drug targets ultimately proposing new possible inhibitors against them. As a matter of fact this study is based on the Interolog method which implies that homologous proteins retain their ability to interact. A distant homolog approach based on Interolog method was employed to speculate MRSA protein homologs in Humans using PSI-BLAST. In addition the protein interaction partners of these homologs as listed in Database of Interacting Proteins (DIP) were predicted to interact with MRSA as well. Moreover, a direct approach using BLAST was also applied so as to attain further confidence in the strategy. Consequently, the common HP-PPIs predicted by both approaches are suggested as potential drug targets (22%) whereas, the unique HP-PPIs estimated only through distant homolog approach are presented as novel drug targets (12%). Furthermore, the most repeated entry in our results was found to be MRSA Histone Deacetylase (HDAC) which was then modeled using SWISS-MODEL. Eventually, small molecules from ZINC, selected randomly, were docked against HDAC using Auto Dock and are suggested as potential binders (inhibitors) based on their energetic profiles. Thus the current study provides basis for further in-depth analysis of such data which not only include MRSA but other deadly pathogens as well.

G4IPDB: A database for G-quadruplex structure forming nucleic acid interacting proteins

Nucleic acid G-quadruplex structure (G4) Interacting Proteins DataBase (G4IPDB) is an important database that contains detailed information about proteins interacting with nucleic acids that forms G-quadruplex structures. G4IPDB is the first database that provides comprehensive information about this interaction at a single platform. This database contains more than 200 entries with details of interaction such as interacting protein name and their synonyms, their UniProt-ID, source organism, target name and its sequences, ∆Tm, binding/dissociation constants, protein gene name, protein FASTA sequence, interacting residue in protein, related PDB entries, interaction ID, graphical view, PMID, author’s name and techniques that were used to detect their interactions. G4IPDB also provides an efficient web-based “G-quadruplex predictor tool” that searches putative G-quadruplex forming sequences simultaneously in both sense and anti-sense strands of the query nucleotide sequence and provides the predicted G score. Studying the interaction between proteins and nucleic acids forming G-quadruplex structures could be of therapeutic significance for various diseases including cancer and neurological disease, therefore, having detail information about their interactions on a single platform would be helpful for the discovery and development of novel therapeutics. G4IPDB can be routinely updated (twice in year) and freely available on http://bsbe.iiti.ac.in/bsbe/ipdb/index.php.

Celecoxib exhibits an anti-gastric cancer effect by targeting focal adhesion and leukocyte transendothelial migration-associated genes

Gastric cancer (GC) is a prevalent cancer, which remains incurable, and therefore requires an alternative treatment method. Celecoxib is a nonsteroidal anti-inflammatory drug that targets cyclooxygenase-2, and exhibits anticancer effects. The present study aimed to investigate the anti-GC mechanism of celecoxib using bioinformatics methods. Gene expression datasets GSE56807 (GC tissues and normal gastric tissues) and GSE54657 (celecoxib-treated and non-treated human GC epithelial AGS cells) were downloaded from the Gene Expression Omnibus database. Two groups of differentially expressed genes (DEGs) were identified using limma package in R language. The criterion for GSE56807 was a false discovery rate of <0.05, while that for GSE54657 was P<0.01. Overlapping DEGs from the two datasets were screened out. Subsequently, pathway enrichment analysis was performed using Database for Annotation, Visualization and Integrated Discovery software (P<0.1; gene count ≥2). In addition, the protein-protein interactions (PPIs) among the overlapped DEGs were obtained based on IntAct, Database of Interacting Proteins, Biomolecular Interaction Network Database and Human Protein Reference Database. Finally, a PPI network was visualized using Cytoscape software. A total of 137 overlapped DEGs were obtained, and DEGs with opposite regulation directions in the two datasets were significantly enriched in focal adhesion and leukocyte transendothelial migration. Subsequently, a PPI network of overlapped DEGs was constructed. Comprehensively, a total of 8 key DEGs [cysteine and glycine rich protein 1 (CSRP1), thrombospondin 1 (THBS1), myosin light chain 9 (MYL9), filamin A (FLNA), actinin alpha 1 (ACTN1), vinculin (VCL), laminin subunit gamma 2 (LAMC2) and claudin 1 (CLDN1)] were upregulated in GC tissues and downregulated in celecoxib-treated cells. In conclusion, celecoxib may exhibit anti-GC effects by suppressing the expression of CSRP1, THBS1, MYL9, FLNA, ACTN1, VCL, LAMC2 and CLDN1, and inhibiting leukocyte transendothelial migration and focal adhesion. However, relevant experiments are required to confirm the conclusion of the present study.

Large-scale extraction of gene interactions from full-text literature using DeepDive.

Motivation: A complete repository of gene–gene interactions is key for understanding cellular processes, human disease and drug response. These gene–gene interactions include both protein–protein interactions and transcription factor interactions. The majority of known interactions are found in the biomedical literature. Interaction databases, such as BioGRID and ChEA, annotate these gene–gene interactions; however, curation becomes difficult as the literature grows exponentially. DeepDive is a trained system for extracting information from a variety of sources, including text. In this work, we used DeepDive to extract both protein–protein and transcription factor interactions from over 100 000 full-text PLOS articles.Methods: We built an extractor for gene–gene interactions that identified candidate gene–gene relations within an input sentence. For each candidate relation, DeepDive computed a probability that the relation was a correct interaction. We evaluated this system against the Database of Interacting Proteins and against randomly curated extractions.Results: Our system achieved 76% precision and 49% recall in extracting direct and indirect interactions involving gene symbols co-occurring in a sentence. For randomly curated extractions, the system achieved between 62% and 83% precision based on direct or indirect interactions, as well as sentence-level and document-level precision. Overall, our system extracted 3356 unique gene pairs using 724 features from over 100 000 full-text articles.Availability and implementation: Application source code is publicly available at https://github.com/edoughty/deepdive_genegene_appContact: russ.altman@stanford.eduSupplementary information: Supplementary data are available at Bioinformatics online.

Tyrosine Kinase Ligand-Receptor Pair Prediction by Using Support Vector Machine.

Receptor tyrosine kinases are essential proteins involved in cellular differentiation and proliferation in vivo and are heavily involved in allergic diseases, diabetes, and onset/proliferation of cancerous cells. Identifying the interacting partner of this protein, a growth factor ligand, will provide a deeper understanding of cellular proliferation/differentiation and other cell processes. In this study, we developed a method for predicting tyrosine kinase ligand-receptor pairs from their amino acid sequences. We collected tyrosine kinase ligand-receptor pairs from the Database of Interacting Proteins (DIP) and UniProtKB, filtered them by removing sequence redundancy, and used them as a dataset for machine learning and assessment of predictive performance. Our prediction method is based on support vector machines (SVMs), and we evaluated several input features suitable for tyrosine kinase for machine learning and compared and analyzed the results. Using sequence pattern information and domain information extracted from sequences as input features, we obtained 0.996 of the area under the receiver operating characteristic curve. This accuracy is higher than that obtained from general protein-protein interaction pair predictions.

Protein-Protein Interactions Prediction Based on Iterative Clique Extension with Gene Ontology Filtering

Cliques (maximal complete subnets) in protein-protein interaction (PPI) network are an important resource used to analyze protein complexes and functional modules. Clique-based methods of predicting PPI complement the data defection from biological experiments. However, clique-based predicting methods only depend on the topology of network. The false-positive and false-negative interactions in a network usually interfere with prediction. Therefore, we propose a method combining clique-based method of prediction and gene ontology (GO) annotations to overcome the shortcoming and improve the accuracy of predictions. According to different GO correcting rules, we generate two predicted interaction sets which guarantee the quality and quantity of predicted protein interactions. The proposed method is applied to the PPI network from the Database of Interacting Proteins (DIP) and most of the predicted interactions are verified by another biological database, BioGRID. The predicted protein interactions are appended to the original protein network, which leads to clique extension and shows the significance of biological meaning.

PRIMOS: An Integrated Database of Reassessed Protein–Protein Interactions Providing Web-Based Access toIn SilicoValidation of Experimentally Derived Data

Steady improvements in proteomics present a bioinformatic challenge to retrieve, store, and process the accumulating and often redundant amount of information. In particular, a large-scale comparison and analysis of protein-protein interaction (PPI) data requires tools for data interpretation as well as validation. At this juncture, the Protein Interaction and Molecule Search (PRIMOS) platform represents a novel web portal that unifies six primary PPI databases (BIND, Biomolecular Interaction Network Database; DIP, Database of Interacting Proteins; HPRD, Human Protein Reference Database; IntAct; MINT, Molecular Interaction Database; and MIPS, Munich Information Center for Protein Sequences) into a single consistent repository, which currently includes more than 196,700 redundancy-removed PPIs. PRIMOS supports three advanced search strategies centering on disease-relevant PPIs, on inter- and intra-organismal crosstalk relations (e.g., pathogen-host interactions), and on highly connected protein nodes analysis ("hub" identification). The main novelties distinguishing PRIMOS from other secondary PPI databases are the reassessment of known PPIs, and the capacity to validate personal experimental data by our peer-reviewed, homology-based validation. This article focuses on definite PRIMOS use cases (presentation of embedded biological concepts, example applications) to demonstrate its broad functionality and practical value. PRIMOS is publicly available at http://primos.fh-hagenberg.at.

Detecting overlapping protein complexes in PPI networks based on robustness

BackgroundRecently, large data sets of protein-protein interactions (PPI) which can be modeled as PPI networks are generated through high-throughput methods. And locally dense regions in PPI networks are very likely to be protein complexes. Since protein complexes play a key role in many biological processes, detecting protein complexes in PPI networks is one of important tasks in post-genomic era. However, PPI networks are often incomplete and noisy, which builds barriers to mining protein complexes.ResultsWe propose a new and effective algorithm based on robustness to detect overlapping clusters as protein complexes in PPI networks. And in order to improve the accuracy of resulting clusters, our algorithm tries to reduce bad effects brought by noise in PPI networks. And in our algorithm, each new cluster begins from a seed and is expanded through adding qualified nodes from the cluster's neighbourhood nodes. Besides, in our algorithm, a new distance measurement method between a cluster K and a node in the neighbours of K is proposed as well. The performance of our algorithm is evaluated by applying it on two PPI networks which are Gavin network and Database of Interacting Proteins (DIP). The results show that our algorithm is better than Markov clustering algorithm (MCL), Clique Percolation method (CPM) and core-attachment based method (CoAch) in terms of F-measure, co-localization and Gene Ontology (GO) semantic similarity.ConclusionsOur algorithm detects locally dense regions or clusters as protein complexes. The results show that protein complexes generated by our algorithm have better quality than those generated by some previous classic methods. Therefore, our algorithm is effective and useful.

Granular support vector machine based on mixed measure

This paper presents a granular support vector machine learning model based on mixed measure, namely M_GSVM, to solve the model error problem produced by mapping, simplifying, granulating or substituting of data for traditional granular support vector machines (GSVM). For M_GSVM, the original data will be mapped into the high-dimensional space by mercer kernel. Then, the data are divided into some granules, and those mixed granules including more information are extracted and trained by support vector machine (SVM). Finally, the decision hyperplane will be corrected through geometric analyzing to reduced model error effectively. The experiment results on UCI benchmark datasets and Interacting Proteins database demonstrate that the proposed M_GSVM model can improve the generalization performance greatly with high learning efficiency synchronously.

ReLiance: a machine learning and literature-based prioritization of receptor—ligand pairings

Motivation: The prediction of receptor—ligand pairings is an important area of research as intercellular communications are mediated by the successful interaction of these key proteins. As the exhaustive assaying of receptor—ligand pairs is impractical, a computational approach to predict pairings is necessary. We propose a workflow to carry out this interaction prediction task, using a text mining approach in conjunction with a state of the art prediction method, as well as a widely accessible and comprehensive dataset.Among several modern classifiers, random forests have been found to be the best at this prediction task. The training of this classifier was carried out using an experimentally validated dataset of Database of Ligand-Receptor Partners (DLRP) receptor—ligand pairs. New examples, co-cited with the training receptors and ligands, are then classified using the trained classifier. After applying our method, we find that we are able to successfully predict receptor—ligand pairs within the GPCR family with a balanced accuracy of 0.96. Upon further inspection, we find several supported interactions that were not present in the Database of Interacting Proteins (DIPdatabase).We have measured the balanced accuracy of our method resulting in high quality predictions stored in the available database ReLiance.Availability: http://homes.esat.kuleuven.be/~bioiuser/ReLianceDB/index.phpContact: yves.moreau@esat.kuleuven.be; ernesto.iacucci@gmail.comSupplementary information: Supplementary data are available at Bioinformatics online.

Functional annotation from the genome sequence of the giant panda

The giant panda is one of the most critically endangered species due to the fragmentation and loss of its habitat. Studying the functions of proteins in this animal, especially specific trait-related proteins, is therefore necessary to protect the species. In this work, the functions of these proteins were investigated using the genome sequence of the giant panda. Data on 21,001 proteins and their functions were stored in the Giant Panda Protein Database, in which the proteins were divided into two groups: 20,179 proteins whose functions can be predicted by GeneScan formed the known-function group, whereas 822 proteins whose functions cannot be predicted by GeneScan comprised the unknown-function group. For the known-function group, we further classified the proteins by molecular function, biological process, cellular component, and tissue specificity. For the unknown-function group, we developed a strategy in which the proteins were filtered by cross-Blast to identify panda-specific proteins under the assumption that proteins related to the panda-specific traits in the unknown-function group exist. After this filtering procedure, we identified 32 proteins (2 of which are membrane proteins) specific to the giant panda genome as compared against the dog and horse genomes. Based on their amino acid sequences, these 32 proteins were further analyzed by functional classification using SVM-Prot, motif prediction using MyHits, and interacting protein prediction using the Database of Interacting Proteins. Nineteen proteins were predicted to be zinc-binding proteins, thus affecting the activities of nucleic acids. The 32 panda-specific proteins will be further investigated by structural and functional analysis.

Using ensemble methods to deal with imbalanced data in predicting protein–protein interactions

In proteins, the number of interacting pairs is usually much smaller than the number of non-interacting ones. So the imbalanced data problem will arise in the field of protein–protein interactions (PPIs) prediction. In this article, we introduce two ensemble methods to solve the imbalanced data problem. These ensemble methods combine the based-cluster under-sampling technique and the fusion classifiers. And then we evaluate the ensemble methods using a dataset from Database of Interacting Proteins (DIP) with 10-fold cross validation. All the prediction models achieve area under the receiver operating characteristic curve (AUC) value about 95%. Our results show that the ensemble classifiers are quite effective in predicting PPIs; we also gain some valuable conclusions on the performance of ensemble methods for PPIs in imbalanced data. The prediction software and all dataset employed in the work can be obtained for free at http://cic.scu.edu.cn/bioinformatics/Ensemble_PPIs/index.html.

PPI_SVM: prediction of protein-protein interactions using machine learning, domain-domain affinities and frequency tables.

Protein-protein interactions (PPI) control most of the biological processes in a living cell. In order to fully understand protein functions, a knowledge of protein-protein interactions is necessary. Prediction of PPI is challenging, especially when the three-dimensional structure of interacting partners is not known. Recently, a novel prediction method was proposed by exploiting physical interactions of constituent domains. We propose here a novel knowledge-based prediction method, namely PPI_SVM, which predicts interactions between two protein sequences by exploiting their domain information. We trained a two-class support vector machine on the benchmarking set of pairs of interacting proteins extracted from the Database of Interacting Proteins (DIP). The method considers all possible combinations of constituent domains between two protein sequences, unlike most of the existing approaches. Moreover, it deals with both single-domain proteins and multi domain proteins; therefore it can be applied to the whole proteome in high-throughput studies. Our machine learning classifier, following a brainstorming approach, achieves accuracy of 86%, with specificity of 95%, and sensitivity of 75%, which are better results than most previous methods that sacrifice recall values in order to boost the overall precision. Our method has on average better sensitivity combined with good selectivity on the benchmarking dataset. The PPI_SVM source code, train/test datasets and supplementary files are available freely in the public domain at: http://code.google.com/p/cmater-bioinfo/. Electronic Supplementary MaterialThe online version of this article (doi: 10.2478/s11658-011-0008-x contains supplementary material, which is available to authorized users.