Protein-protein interaction inference based on semantic similarity of Gene Ontology terms

Explaining protein-protein interactions with knowledge graph-based semantic similarity

The application of artificial intelligence and machine learning methods for several biomedical applications, such as protein-protein interaction prediction, has gained significant traction in recent decades. However, explainability is a key aspect of using machine learning as a tool for scientific discovery. Explainable artificial intelligence approaches help clarify algorithmic mechanisms and identify potential bias in the data. Given the complexity of the biomedical domain, explanations should be grounded in domain knowledge which can be achieved by using ontologies and knowledge graphs. These knowledge graphs express knowledge about a domain by capturing different perspectives of the representation of real-world entities. However, the most popular way to explore knowledge graphs with machine learning is through using embeddings, which are not explainable. As an alternative, knowledge graph-based semantic similarity offers the advantage of being explainable. Additionally, similarity can be computed to capture different semantic aspects within the knowledge graph and increasing the explainability of predictive approaches. We propose a novel method to generate explainable vector representations, KGsim2vec, that uses aspect-oriented semantic similarity features to represent pairs of entities in a knowledge graph. Our approach employs a set of machine learning models, including decision trees, genetic programming, random forest and eXtreme gradient boosting, to predict relations between entities. The experiments reveal that considering multiple semantic aspects when representing the similarity between two entities improves explainability and predictive performance. KGsim2vec performs better than black-box methods based on knowledge graph embeddings or graph neural networks. Moreover, KGsim2vec produces global models that can capture biological phenomena and elucidate data biases.

MM-StackEns: A new deep multimodal stacked generalization approach for protein-protein interaction prediction

Accurate in-silico identification of protein-protein interactions (PPIs) is a long-standing problem in biology, with important implications in protein function prediction and drug design. Current computational approaches predominantly use a single data modality for describing protein pairs, which may not fully capture the characteristics relevant for identifying PPIs. Another limitation of existing methods is their poor generalization to proteins outside the training graph. In this paper, we aim to address these shortcomings by proposing a new ensemble approach for PPI prediction, which learns information from two modalities, corresponding to pairs of sequences and to the graph formed by the training proteins and their interactions. Our approach uses a siamese neural network to process sequence information, while graph attention networks are employed for the network view. For capturing the relationships between the proteins in a pair, we design a new feature fusion module, based on computing the distance between the distributions corresponding to the two proteins. The prediction is made using a stacked generalization procedure, in which the final classifier is represented by a Logistic Regression model trained on the scores predicted by the sequence and graph models. Additionally, we show that protein sequence embeddings obtained using pretrained language models can significantly improve the generalization of PPI methods. The experimental results demonstrate the good performance of our approach, which surpasses all the related work on two Yeast data sets, while outperforming the majority of literature approaches on two Human data sets and on independent multi-species data sets.

Exploration of Computational Aids for Effective Drug Designing and Management of Viral Diseases: A Comprehensive Review

BACKGROUND

Microbial diseases, specifically originating from viruses are the major cause of human mortality all over the world. The current COVID-19 pandemic is a case in point, where the dynamics of the viral-human interactions are still not completely understood, making its treatment a case of trial and error. Scientists are struggling to devise a strategy to contain the pandemic for over a year and this brings to light the lack of understanding of how the virus grows and multiplies in the human body.

METHODS

This paper presents the perspective of the authors on the applicability of computational tools for deep learning and understanding of host-microbe interaction, disease progression and management, drug resistance and immune modulation through in silico methodologies which can aid in effective and selective drug development. The paper has summarized advances in the last five years. The studies published and indexed in leading databases have been included in the review.

RESULTS

Computational systems biology works on an interface of biology and mathematics and intends to unravel the complex mechanisms between the biological systems and the inter and intra species dynamics using computational tools, and high-throughput technologies developed on algorithms, networks and complex connections to simulate cellular biological processes.

CONCLUSION

Computational strategies and modelling integrate and prioritize microbial-host interactions and may predict the conditions in which the fine-tuning attenuates. These microbial-host interactions and working mechanisms are important from the aspect of effective drug designing and fine- tuning the therapeutic interventions.

Analyzing Effect of Multi-Modality in Predicting Protein-Protein Interactions

Nowadays, multiple sources of information about proteins are available such as protein sequences, 3D structures, Gene Ontology (GO), etc. Most of the works on protein-protein interaction (PPI) identification had utilized these information about proteins, mainly sequence-based, but individually. The new advances in deep learning techniques allow us to leverage multiple sources/modalities of proteins, which complement each other. Some recent works have shown that multi-modal PPI models perform better than uni-modal approaches. This paper aims to investigate whether the performance of multi-modal PPI models is always consistent or depends on other factors such as dataset distribution, algorithms used to learn features, etc. We have used three modalities for this study: Protein sequence, 3D structure, and GO. Various techniques, including deep learning algorithms, are employed to extract features from multiple sources of proteins. These feature vectors from different modalities are then integrated in several combinations (bi-modal and tri-modal) to predict PPI. To conduct this study, we have used Human and S. cerevisiae PPI datasets. The obtained results demonstrate the potentiality of a multi-modal approach and deep learning techniques in predicting protein interactions. However, the predictive capability of a model for PPI depends on feature extraction methods as well. Also, increasing the modality does not always ensure performance improvement. In this study, the PPI model integrating two modalities outperforms the designed uni-modal and tri-modal PPI models.

HPIPred: Host-pathogen interactome prediction with phenotypic scoring

Protein-protein interactions (PPIs) are involved in most cellular processes. Unfortunately, current knowledge of host-pathogen interactomes is still very limited. Experimental methods used to detect PPIs have several limitations, including increasing complexity and economic cost in large-scale screenings. Hence, computational methods are commonly used to support experimental data, although they generally suffer from high false-positive rates. To address this issue, we have created HPIPred, a host-pathogen PPI prediction tool based on numerical encoding of physicochemical properties. Unlike other available methods, HPIPred integrates phenotypic data to prioritize biologically meaningful results. We used HPIPred to screen the entire Homo sapiens and Pseudomonas aeruginosa PAO1 proteomes to generate a host-pathogen interactome with 763 interactions displaying a highly connected network topology. Our predictive model can be used to prioritize protein-protein interactions as potential targets for antibacterial drug development. Available at: https://github.com/SysBioUAB/hpi_predictor.

A novel Ontology-guided Attribute Partitioning ensemble learning model for early prediction of cognitive deficits using quantitative Structural MRI in very preterm infants

Structural magnetic resonance imaging studies have shown that brain anatomical abnormalities are associated with cognitive deficits in preterm infants. Brain maturation and geometric features can be used with machine learning models for predicting later neurodevelopmental deficits. However, traditional machine learning models would suffer from a large feature-to-instance ratio (i.e., a large number of features but a small number of instances/samples). Ensemble learning is a paradigm that strategically generates and integrates a library of machine learning classifiers and has been successfully used on a wide variety of predictive modeling problems to boost model performance. Attribute (i.e., feature) bagging method is the most commonly used feature partitioning scheme, which randomly and repeatedly draws feature subsets from the entire feature set. Although attribute bagging method can effectively reduce feature dimensionality to handle the large feature-to-instance ratio, it lacks consideration of domain knowledge and latent relationship among features. In this study, we proposed a novel Ontology-guided Attribute Partitioning (OAP) method to better draw feature subsets by considering the domain-specific relationship among features. With the better-partitioned feature subsets, we developed an ensemble learning framework, which is referred to as OAP-Ensemble Learning (OAP-EL). We applied the OAP-EL to predict cognitive deficits at 2 years of age using quantitative brain maturation and geometric features obtained at term equivalent age in very preterm infants. We demonstrated that the proposed OAP-EL approach significantly outperformed the peer ensemble learning and traditional machine learning approaches.

protein2vec: Predicting Protein-Protein Interactions Based on LSTM

The semantic similarity of gene ontology (GO) terms is widely used to predict protein-protein interactions (PPIs). The traditional semantic similarity measures are based mainly on manually crafted features, which may ignore some important hidden information of the gene ontology. Moreover, those methods usually obtain the similarity between proteins from similarity between GO terms by some simple statistical rules, such as MAX and BMA (best-match average), oversimplifying the possible complex relationship between the proteins and the GO terms annotated with them. To overcome the two deficiencies, we propose a new method named protein2vec, which characterizes a protein with a vector based on the GO terms annotated to it and combines the information of both the GO and known PPIs. We firstly try to apply the network embedding algorithm on the GO network to generate feature vectors for each GO term. Then, Long Short-Time Memory (LSTM) encodes the feature vectors of the GO terms annotated with a protein into another vector (called protein vector). Finally, two protein vectors are forwarded into a feedforward neural network to predict the interaction between the two corresponding proteins. The experimental results show that protein2vec outperforms almost all commonly used traditional semantic similarity methods.

PFmulDL: a novel strategy enabling multi-class and multi-label protein function annotation by integrating diverse deep learning methods

Bioinformatic annotation of protein function is essential but extremely sophisticated, which asks for extensive efforts to develop effective prediction method. However, the existing methods tend to amplify the representativeness of the families with large number of proteins by misclassifying the proteins in the families with small number of proteins. That is to say, the ability of the existing methods to annotate proteins in the 'rare classes' remains limited. Herein, a new protein function annotation strategy, PFmulDL, integrating multiple deep learning methods, was thus constructed. First, the recurrent neural network was integrated, for the first time, with the convolutional neural network to facilitate the function annotation. Second, a transfer learning method was introduced to the model construction for further improving the prediction performances. Third, based on the latest data of Gene Ontology, the newly constructed model could annotate the largest number of protein families comparing with the existing methods. Finally, this newly constructed model was found capable of significantly elevating the prediction performance for the 'rare classes' without sacrificing that for the 'major classes'. All in all, due to the emerging requirements on improving the prediction performance for the proteins in 'rare classes', this new strategy would become an essential complement to the existing methods for protein function prediction. All the models and source codes are freely available and open to all users at: https://github.com/idrblab/PFmulDL.

TransformerGO: predicting protein-protein interactions by modelling the attention between sets of gene ontology terms

MOTIVATION

Protein-protein interactions (PPIs) play a key role in diverse biological processes but only a small subset of the interactions has been experimentally identified. Additionally, high-throughput experimental techniques that detect PPIs are known to suffer various limitations, such as exaggerated false positives and negatives rates. The semantic similarity derived from the Gene Ontology (GO) annotation is regarded as one of the most powerful indicators for protein interactions. However, while computational approaches for prediction of PPIs have gained popularity in recent years, most methods fail to capture the specificity of GO terms.

RESULTS

We propose TransformerGO, a model that is capable of capturing the semantic similarity between GO sets dynamically using an attention mechanism. We generate dense graph embeddings for GO terms using an algorithmic framework for learning continuous representations of nodes in networks called node2vec. TransformerGO learns deep semantic relations between annotated terms and can distinguish between negative and positive interactions with high accuracy. TransformerGO outperforms classic semantic similarity measures on gold standard PPI datasets and state-of-the-art machine-learning-based approaches on large datasets from Saccharomyces cerevisiae and Homo sapiens. We show how the neural attention mechanism embedded in the transformer architecture detects relevant functional terms when predicting interactions.

AVAILABILITY AND IMPLEMENTATION

https://github.com/Ieremie/TransformerGO.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Evaluating semantic similarity methods for comparison of text-derived phenotype profiles

Background

Semantic similarity is a valuable tool for analysis in biomedicine. When applied to phenotype profiles derived from clinical text, they have the capacity to enable and enhance ‘patient-like me’ analyses, automated coding, differential diagnosis, and outcome prediction. While a large body of work exists exploring the use of semantic similarity for multiple tasks, including protein interaction prediction, and rare disease differential diagnosis, there is less work exploring comparison of patient phenotype profiles for clinical tasks. Moreover, there are no experimental explorations of optimal parameters or better methods in the area.

Methods

We develop a platform for reproducible benchmarking and comparison of experimental conditions for patient phentoype similarity. Using the platform, we evaluate the task of ranking shared primary diagnosis from uncurated phenotype profiles derived from all text narrative associated with admissions in the medical information mart for intensive care (MIMIC-III).

Results

300 semantic similarity configurations were evaluated, as well as one embedding-based approach. On average, measures that did not make use of an external information content measure performed slightly better, however the best-performing configurations when measured by area under receiver operating characteristic curve and Top Ten Accuracy used term-specificity and annotation-frequency measures.

Conclusion

We identified and interpreted the performance of a large number of semantic similarity configurations for the task of classifying diagnosis from text-derived phenotype profiles in one setting. We also provided a basis for further research on other settings and related tasks in the area.

OUP accepted manuscript

Motivation

There is a plethora of measures to evaluate functional similarity (FS) of genes based on their co-expression, protein–protein interactions and sequence similarity. These measures are typically derived from hand-engineered and application-specific metrics to quantify the degree of shared information between two genes using their Gene Ontology (GO) annotations.

Results

We introduce deepSimDEF, a deep learning method to automatically learn FS estimation of gene pairs given a set of genes and their GO annotations. deepSimDEF’s key novelty is its ability to learn low-dimensional embedding vector representations of GO terms and gene products and then calculate FS using these learned vectors. We show that deepSimDEF can predict the FS of new genes using their annotations: it outperformed all other FS measures by >5–10% on yeast and human reference datasets on protein–protein interactions, gene co-expression and sequence homology tasks. Thus, deepSimDEF offers a powerful and adaptable deep neural architecture that can benefit a wide range of problems in genomics and proteomics, and its architecture is flexible enough to support its extension to any organism.

Availability and implementation

Source code and data are available at https://github.com/ahmadpgh/deepSimDEF

Supplementary information

Supplementary data are available at Bioinformatics online.

A New Family of Similarity Measures for Scoring Confidence of Protein Interactions Using Gene Ontology

The large-scale protein-protein interaction (PPI) data has the potential to play a significant role in the endeavor of understanding cellular processes. However, the presence of a considerable fraction of false positives is a bottleneck in realizing this potential. There have been continuous efforts to utilize complementary resources for scoring confidence of PPIs in a manner that false positive interactions get a low confidence score. Gene Ontology (GO), a taxonomy of biological terms to represent the properties of gene products and their relations, has been widely used for this purpose. We utilize GO to introduce a new set of specificity measures: Relative Depth Specificity (RDS), Relative Node-based Specificity (RNS), and Relative Edge-based Specificity (RES), leading to a new family of similarity measures. We use these similarity measures to obtain a confidence score for each PPI. We evaluate the new measures using four different benchmarks. We show that all the three measures are quite effective. Notably, RNS and RES more effectively distinguish true PPIs from false positives than the existing alternatives. RES also shows a robust set-discriminating power and can be useful for protein functional clustering as well.

Deep Neural Network and Extreme Gradient Boosting Based Hybrid Classifier for Improved Prediction of Protein-Protein Interaction

Understanding the behavioral process of life and disease-causing mechanism, knowledge regarding protein-protein interactions (PPI) is essential. In this paper, a novel hybrid approach combining deep neural network (DNN) and extreme gradient boosting classifier (XGB) is employed for predicting PPI. The hybrid classifier (DNN-XGB) uses a fusion of three sequence-based features, amino acid composition (AAC), conjoint triad composition (CT), and local descriptor (LD) as inputs. The DNN extracts the hidden information through a layer-wise abstraction from the raw features that are passed through the XGB classifier. The 5-fold cross-validation accuracy for intraspecies interactions dataset of Saccharomyces cerevisiae (core subset), Helicobacter pylori, Saccharomyces cerevisiae, and Human are 98.35, 96.19, 97.37, and 99.74 percent respectively. Similarly, accuracies of 98.50 and 97.25 percent are achieved for interspecies interaction dataset of Human- Bacillus Anthracis and Human- Yersinia pestis datasets, respectively. The improved prediction accuracies obtained on the independent test sets and network datasets indicate that the DNN-XGB can be used to predict cross-species interactions. It can also provide new insights into signaling pathway analysis, predicting drug targets, and understanding disease pathogenesis. Improved performance of the proposed method suggests that the hybrid classifier can be used as a useful tool for PPI prediction. The datasets and source codes are available at: https://github.com/SatyajitECE/DNN-XGB-for-PPI-Prediction.

Effects of Negation and Uncertainty Stratification on Text-Derived Patient Profile Similarity

Semantic similarity is a useful approach for comparing patient phenotypes, and holds the potential of an effective method for exploiting text-derived phenotypes for differential diagnosis, text and document classification, and outcome prediction. While approaches for context disambiguation are commonly used in text mining applications, forming a standard component of information extraction pipelines, their effects on semantic similarity calculations have not been widely explored. In this work, we evaluate how inclusion and disclusion of negated and uncertain mentions of concepts from text-derived phenotypes affects similarity of patients, and the use of those profiles to predict diagnosis. We report on the effectiveness of these approaches and report a very small, yet significant, improvement in performance when classifying primary diagnosis over MIMIC-III patient visits.

Benchmark Evaluation of Protein-Protein Interaction Prediction Algorithms

Protein-protein interactions (PPIs) perform various functions and regulate processes throughout cells. Knowledge of the full network of PPIs is vital to biomedical research, but most of the PPIs are still unknown. As it is infeasible to discover all of them experimentally due to technical and resource limitations, computational prediction of PPIs is essential and accurately assessing the performance of algorithms is required before further application or translation. However, many published methods compose their evaluation datasets incorrectly, using a higher proportion of positive class data than occuring naturally, leading to exaggerated performance. We re-implemented various published algorithms and evaluated them on datasets with realistic data compositions and found that their performance is overstated in original publications; with several methods outperformed by our control models built on 'illogical' and random number features. We conclude that these methods are influenced by an over-characterization of some proteins in the literature and due to scale-free nature of PPI network and that they fail when tested on all possible protein pairs. Additionally, we found that sequence-only-based algorithms performed worse than those that employ functional and expression features. We present a benchmark evaluation of many published algorithms for PPI prediction. The source code of our implementations and the benchmark datasets created here are made available in open source.

Semantic similarity and machine learning with ontologies

Ontologies have long been employed in the life sciences to formally represent and reason over domain knowledge and they are employed in almost every major biological database. Recently, ontologies are increasingly being used to provide background knowledge in similarity-based analysis and machine learning models. The methods employed to combine ontologies and machine learning are still novel and actively being developed. We provide an overview over the methods that use ontologies to compute similarity and incorporate them in machine learning methods; in particular, we outline how semantic similarity measures and ontology embeddings can exploit the background knowledge in ontologies and how ontologies can provide constraints that improve machine learning models. The methods and experiments we describe are available as a set of executable notebooks, and we also provide a set of slides and additional resources at https://github.com/bio-ontology-research-group/machine-learning-with-ontologies.

Approximate search for known gene clusters in new genomes using PQ-trees

Gene clusters are groups of genes that are co-locally conserved across various genomes, not necessarily in the same order. Their discovery and analysis is valuable in tasks such as gene annotation and prediction of gene interactions, and in the study of genome organization and evolution. The discovery of conserved gene clusters in a given set of genomes is a well studied problem, but with the rapid sequencing of prokaryotic genomes a new problem is inspired. Namely, given an already known gene cluster that was discovered and studied in one genomic dataset, to identify all the instances of the gene cluster in a given new genomic sequence. Thus, we define a new problem in comparative genomics, denoted PQ-Tree Search that takes as input a PQ-tree T representing the known gene orders of a gene cluster of interest, a gene-to-gene substitution scoring function h, integer arguments \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$d_T$$\end{document}dT and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$d_S$$\end{document}dS, and a new sequence of genes S. The objective is to identify in S approximate new instances of the gene cluster; These instances could vary from the known gene orders by genome rearrangements that are constrained by T, by gene substitutions that are governed by h, and by gene deletions and insertions that are bounded from above by \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$d_T$$\end{document}dT and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$d_S$$\end{document}dS, respectively. We prove that PQ-Tree Search is NP-hard and propose a parameterized algorithm that solves the optimization variant of PQ-Tree Search in \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O^*(2^{\gamma })$$\end{document}O∗(2γ) time, where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\gamma$$\end{document}γ is the maximum degree of a node in T and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O^*$$\end{document}O∗ is used to hide factors polynomial in the input size. The algorithm is implemented as a search tool, denoted PQFinder, and applied to search for instances of chromosomal gene clusters in plasmids, within a dataset of 1,487 prokaryotic genomes. We report on 29 chromosomal gene clusters that are rearranged in plasmids, where the rearrangements are guided by the corresponding PQ-trees. One of these results, coding for a heavy metal efflux pump, is further analysed to exemplify how PQFinder can be harnessed to reveal interesting new structural variants of known gene clusters.

Protein-Protein Interactions Efficiently Modeled by Residue Cluster Classes

Predicting protein-protein interactions (PPI) represents an important challenge in structural bioinformatics. Current computational methods display different degrees of accuracy when predicting these interactions. Different factors were proposed to help improve these predictions, including choosing the proper descriptors of proteins to represent these interactions, among others. In the current work, we provide a representative protein structure that is amenable to PPI classification using machine learning approaches, referred to as residue cluster classes. Through sampling and optimization, we identified the best algorithm-parameter pair to classify PPI from more than 360 different training sets. We tested these classifiers against PPI datasets that were not included in the training set but shared sequence similarity with proteins in the training set to reproduce the situation of most proteins sharing sequence similarity with others. We identified a model with almost no PPI error (96-99% of correctly classified instances) and showed that residue cluster classes of protein pairs displayed a distinct pattern between positive and negative protein interactions. Our results indicated that residue cluster classes are structural features relevant to model PPI and provide a novel tool to mathematically model the protein structure/function relationship.

Evolving knowledge graph similarity for supervised learning in complex biomedical domains

Background

In recent years, biomedical ontologies have become important for describing existing biological knowledge in the form of knowledge graphs. Data mining approaches that work with knowledge graphs have been proposed, but they are based on vector representations that do not capture the full underlying semantics. An alternative is to use machine learning approaches that explore semantic similarity. However, since ontologies can model multiple perspectives, semantic similarity computations for a given learning task need to be fine-tuned to account for this. Obtaining the best combination of semantic similarity aspects for each learning task is not trivial and typically depends on expert knowledge.

Results

We have developed a novel approach, evoKGsim, that applies Genetic Programming over a set of semantic similarity features, each based on a semantic aspect of the data, to obtain the best combination for a given supervised learning task. The approach was evaluated on several benchmark datasets for protein-protein interaction prediction using the Gene Ontology as the knowledge graph to support semantic similarity, and it outperformed competing strategies, including manually selected combinations of semantic aspects emulating expert knowledge. evoKGsim was also able to learn species-agnostic models with different combinations of species for training and testing, effectively addressing the limitations of predicting protein-protein interactions for species with fewer known interactions.

Conclusions

evoKGsim can overcome one of the limitations in knowledge graph-based semantic similarity applications: the need to expertly select which aspects should be taken into account for a given application. Applying this methodology to protein-protein interaction prediction proved successful, paving the way to broader applications.

A Computational Framework for Predicting Direct Contacts and Substructures within Protein Complexes

Understanding the physical arrangement of subunits within protein complexes potentially provides valuable clues about how the subunits work together and how the complexes function. The majority of recent research focuses on identifying protein complexes as a whole and seldom studies the inner structures within complexes. In this study, we propose a computational framework to predict direct contacts and substructures within protein complexes. In this framework, we first train a supervised learning model of l₂-regularized logistic regression to learn the patterns of direct and indirect interactions within complexes, from where physical subunit interaction networks are predicted. Then, to infer substructures within complexes, we apply a graph clustering method (i.e., maximum modularity clustering (MMC)) and a gene ontology (GO) semantic similarity based functional clustering on partially- and fully-connected networks, respectively. Computational results show that the proposed framework achieves fairly good performance of cross validation and independent test in terms of detecting direct contacts between subunits. Functional analyses further demonstrate the rationality of partitioning the subunits into substructures via the MMC algorithm and functional clustering.

Protein complex identification based on weighted PPI network with multi-source information

Proteins form complexes to accomplish biological functions such as transcription of DNA, translation of mRNA and cell growth. Detection of protein complexes from protein-protein interaction (PPI) networks is the first step for the analysis of biological processes and pathways. Here, we propose a new framework by incorporating Gene Ontology (GO), amino acid background frequency (AABF) and data from von Mering (von Mering data) to identify protein complexes. Firstly, based on the semantic similarity of GO, we construct a weighted PPI network. Secondly, von Mering data is added to construct six types of weighted graphs. Lastly, by integrating density, diameter and cosine similarity, we define a new condition for clustering proteins in these weighted protein network by selecting specific node as key node. Comparison and analysis results indicate that our proposed method could achieve better performances than some classic existing approaches in regard to f-measure and precision.

Computational identification of protein-protein interactions in model plant proteomes

Protein-protein interactions (PPIs) play essential roles in many biological processes. A PPI network provides crucial information on how biological pathways are structured and coordinated from individual protein functions. In the past two decades, large-scale PPI networks of a handful of organisms were determined by experimental techniques. However, these experimental methods are time-consuming, expensive, and are not easy to perform on new target organisms. Large-scale PPI data is particularly sparse in plant organisms. Here, we developed a computational approach for detecting PPIs trained and tested on known PPIs of Arabidopsis thaliana and applied to three plants, Arabidopsis thaliana, Glycine max (soybean), and Zea mays (maize) to discover new PPIs on a genome-scale. Our method considers a variety of features including protein sequences, gene co-expression, functional association, and phylogenetic profiles. This is the first work where a PPI prediction method was developed for is the first PPI prediction method applied on benchmark datasets of Arabidopsis. The method showed a high prediction accuracy of over 90% and very high precision of close to 1.0. We predicted 50,220 PPIs in Arabidopsis thaliana, 13,175,414 PPIs in corn, and 13,527,834 PPIs in soybean. Newly predicted PPIs were classified into three confidence levels according to the availability of existing supporting evidence and discussed. Predicted PPIs in the three plant genomes are made available for future reference.

Multi-domain semantic similarity in biomedical research

BACKGROUND

Given the increasing amount of biomedical resources that are being annotated with concepts from more than one ontology and covering multiple domains of knowledge, it is important to devise mechanisms to compare these resources that take into account the various domains of annotation. For example, metabolic pathways are annotated with their enzymes and their metabolites, and thus similarity measures should compare them with respect to both of those domains simultaneously.

RESULTS

In this paper, we propose two approaches to lift existing single-ontology semantic similarity measures into multi-domain measures. The aggregative approach compares domains independently and averages the various similarity values into a final score. The integrative approach integrates all the relevant ontologies into a single one, calculating similarity in the resulting multi-domain ontology using the single-ontology measure.

CONCLUSIONS

We evaluated the two approaches in a multidisciplinary epidemiology dataset by evaluating the capacity of the similarity measures to predict new annotations based on the existing ones. The results show a promising increase in performance of the multi-domain measures over the single-ontology ones in the vast majority of the cases. These results show that multi-domain measures outperform single-domain ones, and should be considered by the community as a starting point to study more efficient multi-domain semantic similarity measures.

Review of computational methods for virus-host protein interaction prediction: a case study on novel Ebola-human interactions

Identification of potential virus-host interactions is useful and vital to control the highly infectious virus-caused diseases. This may contribute toward development of new drugs to treat the viral infections. Recently, database records of clinically and experimentally validated interactions between a small set of human proteins and Ebola virus (EBOV) have been published. Using the information of the known human interaction partners of EBOV, our main objective is to identify a set of proteins that may interact with EBOV proteins. Here, we first review the state-of-the-art, computational methods used for prediction of novel virus-host interactions for infectious diseases followed by a case study on EBOV-human interactions. The assessment result shows that the predicted human host proteins are highly similar with known human interaction partners of EBOV in the context of structure and semantics and are responsible for similar biochemical activities, pathways and host-pathogen relationships.

Computational Methods for Predicting Protein-Protein Interactions Using Various Protein Features

Understanding protein-protein interactions (PPIs) in a cell is essential for learning protein functions, pathways, and mechanism of diseases. PPIs are also important targets for developing drugs. Experimental methods, both small-scale and large-scale, have identified PPIs in several model organisms. However, results cover only a part of PPIs of organisms; moreover, there are many organisms whose PPIs have not yet been investigated. To complement experimental methods, many computational methods have been developed that predict PPIs from various characteristics of proteins. Here we provide an overview of literature reports to classify computational PPI prediction methods that consider different features of proteins, including protein sequence, genomes, protein structure, function, PPI network topology, and those which integrate multiple methods. © 2018 by John Wiley & Sons, Inc.

Prediction of subcellular location of apoptosis proteins by incorporating PsePSSM and DCCA coefficient based on LFDA dimensionality reduction

BACKGROUND

Apoptosis is associated with some human diseases, including cancer, autoimmune disease, neurodegenerative disease and ischemic damage, etc. Apoptosis proteins subcellular localization information is very important for understanding the mechanism of programmed cell death and the development of drugs. Therefore, the prediction of subcellular localization of apoptosis protein is still a challenging task.

RESULTS

In this paper, we propose a novel method for predicting apoptosis protein subcellular localization, called PsePSSM-DCCA-LFDA. Firstly, the protein sequences are extracted by combining pseudo-position specific scoring matrix (PsePSSM) and detrended cross-correlation analysis coefficient (DCCA coefficient), then the extracted feature information is reduced dimensionality by LFDA (local Fisher discriminant analysis). Finally, the optimal feature vectors are input to the SVM classifier to predict subcellular location of the apoptosis proteins. The overall prediction accuracy of 99.7, 99.6 and 100% are achieved respectively on the three benchmark datasets by the most rigorous jackknife test, which is better than other state-of-the-art methods.

CONCLUSION

The experimental results indicate that our method can significantly improve the prediction accuracy of subcellular localization of apoptosis proteins, which is quite high to be able to become a promising tool for further proteomics studies. The source code and all datasets are available at https://github.com/QUST-BSBRC/PsePSSM-DCCA-LFDA/ .

An improved approach to infer protein-protein interaction based on a hierarchical vector space model

BACKGROUND

Comparing and classifying functions of gene products are important in today's biomedical research. The semantic similarity derived from the Gene Ontology (GO) annotation has been regarded as one of the most widely used indicators for protein interaction. Among the various approaches proposed, those based on the vector space model are relatively simple, but their effectiveness is far from satisfying.

RESULTS

We propose a Hierarchical Vector Space Model (HVSM) for computing semantic similarity between different genes or their products, which enhances the basic vector space model by introducing the relation between GO terms. Besides the directly annotated terms, HVSM also takes their ancestors and descendants related by "is_a" and "part_of" relations into account. Moreover, HVSM introduces the concept of a Certainty Factor to calibrate the semantic similarity based on the number of terms annotated to genes. To assess the performance of our method, we applied HVSM to Homo sapiens and Saccharomyces cerevisiae protein-protein interaction datasets. Compared with TCSS, Resnik, and other classic similarity measures, HVSM achieved significant improvement for distinguishing positive from negative protein interactions. We also tested its correlation with sequence, EC, and Pfam similarity using online tool CESSM.

CONCLUSIONS

HVSM showed an improvement of up to 4% compared to TCSS, 8% compared to IntelliGO, 12% compared to basic VSM, 6% compared to Resnik, 8% compared to Lin, 11% compared to Jiang, 8% compared to Schlicker, and 11% compared to SimGIC using AUC scores. CESSM test showed HVSM was comparable to SimGIC, and superior to all other similarity measures in CESSM as well as TCSS. Supplementary information and the software are available at https://github.com/kejia1215/HVSM .

Complex network theory for the identification and assessment of candidate protein targets

In this work we use complex network theory to provide a statistical model of the connectivity patterns of human proteins and their interaction partners. Our intention is to identify important proteins that may be predisposed to be potential candidates as drug targets for therapeutic interventions. Target proteins usually have more interaction partners than non-target proteins, but there are no hard-and-fast rules for defining the actual number of interactions. We devise a statistical measure for identifying hub proteins, we score our target proteins with gene ontology annotations. The important druggable protein targets are likely to have similar biological functions that can be assessed for their potential therapeutic value. Our system provides a statistical analysis of the local and distant neighborhood protein interactions of the potential targets using complex network measures. This approach builds a more accurate model of drug-to-target activity and therefore the likely impact on treating diseases. We integrate high quality protein interaction data from the HINT database and disease associated proteins from the DrugTarget database. Other sources include biological knowledge from Gene Ontology and drug information from DrugBank. The problem is a very challenging one since the data is highly imbalanced between target proteins and the more numerous nontargets. We use undersampling on the training data and build Random Forest classifier models which are used to identify previously unclassified target proteins. We validate and corroborate these findings from the available literature.

GFD-Net: A novel semantic similarity methodology for the analysis of gene networks

Since the popularization of biological network inference methods, it has become crucial to create methods to validate the resulting models. Here we present GFD-Net, the first methodology that applies the concept of semantic similarity to gene network analysis. GFD-Net combines the concept of semantic similarity with the use of gene network topology to analyze the functional dissimilarity of gene networks based on Gene Ontology (GO). The main innovation of GFD-Net lies in the way that semantic similarity is used to analyze gene networks taking into account the network topology. GFD-Net selects a functionality for each gene (specified by a GO term), weights each edge according to the dissimilarity between the nodes at its ends and calculates a quantitative measure of the network functional dissimilarity, i.e. a quantitative value of the degree of dissimilarity between the connected genes. The robustness of GFD-Net as a gene network validation tool was demonstrated by performing a ROC analysis on several network repositories. Furthermore, a well-known network was analyzed showing that GFD-Net can also be used to infer knowledge. The relevance of GFD-Net becomes more evident in Section "GFD-Net applied to the study of human diseases" where an example of how GFD-Net can be applied to the study of human diseases is presented. GFD-Net is available as an open-source Cytoscape app which offers a user-friendly interface to configure and execute the algorithm as well as the ability to visualize and interact with the results(http://apps.cytoscape.org/apps/gfdnet).