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.     

Multi Label Learning for Prediction of Human Protein Subcellular Localizations

Predicting protein subcellular locations has attracted much attention in the past decade. However, one of the most challenging problems is that many proteins were found simultaneously existing in, or moving between, two or more different cell components in a eukaryotic cell. Seldom previous predictors were able to deal with such multiplex proteins although they have extremely important implications in future drug discovery in terms of their specific subcellular targeting. Approximately 20% of the human proteome consists of such multiplex proteins with multiple sample labels. In order to efficiently handle such multiplex human proteins, we have developed a novel multi-label (ML) learning and prediction framework called ML-PLoc, which decomposes the multi-label prediction problem into multiple independent binary classification problems. ML-PLoc is constructed based on support vector machine (SVM) and sequential evolution information. Experimental results show that ML-PLoc can achieve an overall accuracy 64.6% and recall ratio 67.2% on a benchmark dataset consisting of 14 human subcellular locations, and is very powerful for dealing with multiplex proteins. The current approach represents a new strategy to deal with the multi-label biological problems. ML-PLoc software is freely available for academic use at: http://www.csbio.sjtu.edu.cn/bioinf/ML-PLoc.     

A New Supervised Over-Sampling Algorithm with Application to Protein-Nucleotide Binding Residue Prediction

Protein-nucleotide interactions are ubiquitous in a wide variety of biological processes. Accurately identifying interaction residues solely from protein sequences is useful for both protein function annotation and drug design, especially in the post-genomic era, as large volumes of protein data have not been functionally annotated. Protein-nucleotide binding residue prediction is a typical imbalanced learning problem, where binding residues are extremely fewer in number than non-binding residues. Alleviating the severity of class imbalance has been demonstrated to be a promising means of improving the prediction performance of a machine-learning-based predictor for class imbalance problems. However, little attention has been paid to the negative impact of class imbalance on protein-nucleotide binding residue prediction. In this study, we propose a new supervised over-sampling algorithm that synthesizes additional minority class samples to address class imbalance. The experimental results from protein-nucleotide interaction datasets demonstrate that the proposed supervised over-sampling algorithm can relieve the severity of class imbalance and help to improve prediction performance. Based on the proposed over-sampling algorithm, a predictor, called TargetSOS, is implemented for protein-nucleotide binding residue prediction. Cross-validation tests and independent validation tests demonstrate the effectiveness of TargetSOS. The web-server and datasets used in this study are freely available at http://www.csbio.sjtu.edu.cn/bioinf/TargetSOS/.     

PseAAC-Builder: a cross-platform stand-alone program for generating various special Chou's pseudo-amino acid compositions

The pseudo-amino acid composition has been widely used to convert complicated protein sequences with various lengths to fixed length digital feature vectors while keeping considerable sequence order information. However, so far the only software available to the public is the web server PseAAC (http://www.csbio.sjtu.edu.cn/bioinf/PseAAC), which has some limitations in dealing with large-scale datasets. Here, we propose a new cross-platform stand-alone software program, called PseAAC-Builder (http://www.pseb.sf.net), which can be used to generate various modes of Chou's pseudo-amino acid composition in a much more efficient and flexible way. It is anticipated that PseAAC-Builder may become a useful tool for studying various protein attributes.     

GraphBind: protein structural context embedded rules learned by hierarchical graph neural networks for recognizing nucleic-acid-binding residues

Knowledge of the interactions between proteins and nucleic acids is the basis of understanding various biological activities and designing new drugs. How to accurately identify the nucleic-acid-binding residues remains a challenging task. In this paper, we propose an accurate predictor, GraphBind, for identifying nucleic-acid-binding residues on proteins based on an end-to-end graph neural network. Considering that binding sites often behave in highly conservative patterns on local tertiary structures, we first construct graphs based on the structural contexts of target residues and their spatial neighborhood. Then, hierarchical graph neural networks (HGNNs) are used to embed the latent local patterns of structural and bio-physicochemical characteristics for binding residue recognition. We comprehensively evaluate GraphBind on DNA/RNA benchmark datasets. The results demonstrate the superior performance of GraphBind than state-of-the-art methods. Moreover, GraphBind is extended to other ligand-binding residue prediction to verify its generalization capability. Web server of GraphBind is freely available at http://www.csbio.sjtu.edu.cn/bioinf/GraphBind/.     

PredCSF: an integrated feature-based approach for predicting conotoxin superfamily

Conotoxins are small disulfide-rich peptides that are invaluable channel-targeted peptides and target neuronal receptors. They show prospects for being potent pharmaceuticals in the treatment of Alzheimer's disease, Parkinson's disease, and epilepsy. Accurate and fast prediction of conotoxin superfamily is very helpful towards the understanding of its biological and pharmacological functions especially in the post-genomic era. In the present study, we have developed a novel approach called PredCSF for predicting the conotoxin superfamily from the amino acid sequence directly based on fusing different kinds of sequential features by using modified one-versus-rest SVMs. The input features to the PredCSF classifiers are composed of physicochemical properties, evolutionary information, predicted second structure and amino acid composition, where the most important features are further screened by random forest feature selection to improve the prediction performance. The prediction results show that PredCSF can obtain an overall accuracy of 90.65% based on a benchmark dataset constructed from the most recent database, which consists of 4 main conotoxin superfamilies and 1 class of non-conotoxin class. Systematic experiments also show that combing different features is helpful for enhancing the prediction power when dealing with complex biological problems. PredCSF is expected to be a powerful tool for in silico identification of novel conotonxins and is freely available for academic use at http://www.csbio.sjtu.edu.cn/bioinf/PredCSF.     

Predicting protein fold pattern with functional domain and sequential evolution information

The fold pattern of a protein is one level deeper than its structural classification, and hence is more challenging and complicated for prediction. Many efforts have been made in this regard, but so far all the reported success rates are still under 70%, indicating that it is extremely difficult to enhance the success rate even by 1% or 2%. To address this problem, here a novel approach is proposed that is featured by combining the functional domain information and the sequential evolution information through a fusion ensemble classifier. The predictor thus developed is called PFP-FunDSeqE. Tests were performed for identifying proteins among their 27 fold patterns. Compared with the existing predictors tested by a same stringent benchmark dataset, the new predictor can, for the first time, achieve over 70% success rate. The PFP-FunDSeqE predictor is freely available to the public as a web server at http://www.csbio.sjtu.edu.cn/bioinf/PFP-FunDSeqE/.     

ProtIdent: a web server for identifying proteases and their types by fusing functional domain and sequential evolution information

Proteases are vitally important to life cycles and have become a main target in drug development. According to their action mechanisms, proteases are classified into six types: (1) aspartic, (2) cysteine, (3) glutamic, (4) metallo, (5) serine, and (6) threonine. Given the sequence of an uncharacterized protein, can we identify whether it is a protease or non-protease? If it is, what type does it belong to? To address these problems, a 2-layer predictor, called "ProtIdent", is developed by fusing the functional domain and sequential evolution information: the first layer is for identifying the query protein as protease or non-protease; if it is a protease, the process will automatically go to the second layer to further identify it among the six types. The overall success rates in both cases by rigorous cross-validation tests were higher than 92%. ProtIdent is freely accessible to the public as a web server at http://www.csbio.sjtu.edu.cn/bioinf/Protease.     

Plant-mPLoc: A Top-Down Strategy to Augment the Power for Predicting Plant Protein Subcellular Localization

One of the fundamental goals in proteomics and cell biology is to identify the functions of proteins in various cellular organelles and pathways. Information of subcellular locations of proteins can provide useful insights for revealing their functions and understanding how they interact with each other in cellular network systems. Most of the existing methods in predicting plant protein subcellular localization can only cover three or four location sites, and none of them can be used to deal with multiplex plant proteins that can simultaneously exist at two, or move between, two or more different location sites. Actually, such multiplex proteins might have special biological functions worthy of particular notice. The present study was devoted to improve the existing plant protein subcellular location predictors from the aforementioned two aspects. A new predictor called “Plant-mPLoc” is developed by integrating the gene ontology information, functional domain information, and sequential evolutionary information through three different modes of pseudo amino acid composition. It can be used to identify plant proteins among the following 12 location sites: (1) cell membrane, (2) cell wall, (3) chloroplast, (4) cytoplasm, (5) endoplasmic reticulum, (6) extracellular, (7) Golgi apparatus, (8) mitochondrion, (9) nucleus, (10) peroxisome, (11) plastid, and (12) vacuole. Compared with the existing methods for predicting plant protein subcellular localization, the new predictor is much more powerful and flexible. Particularly, it also has the capacity to deal with multiple-location proteins, which is beyond the reach of any existing predictors specialized for identifying plant protein subcellular localization. As a user-friendly web-server, Plant-mPLoc is freely accessible at http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/. Moreover, for the convenience of the vast majority of experimental scientists, a step-by-step guide is provided on how to use the web-server to get the desired results. It is anticipated that the Plant-mPLoc predictor as presented in this paper will become a very useful tool in plant science as well as all the relevant areas.     

LabCaS: Labeling calpain substrate cleavage sites from amino acid sequence using conditional random fields

The calpain family of Ca²⁺‐dependent cysteine proteases plays a vital role in many important biological processes which is closely related with a variety of pathological states. Activated calpains selectively cleave relevant substrates at specific cleavage sites, yielding multiple fragments that can have different functions from the intact substrate protein. Until now, our knowledge about the calpain functions and their substrate cleavage mechanisms are limited because the experimental determination and validation on calpain binding are usually laborious and expensive. In this work, we aim to develop a new computational approach (LabCaS) for accurate prediction of the calpain substrate cleavage sites from amino acid sequences. To overcome the imbalance of negative and positive samples in the machine‐learning training which have been suffered by most of the former approaches when splitting sequences into short peptides, we designed a conditional random field algorithm that can label the potential cleavage sites directly from the entire sequences. By integrating the multiple amino acid features and those derived from sequences, LabCaS achieves an accurate recognition of the cleave sites for most calpain proteins. In a jackknife test on a set of 129 benchmark proteins, LabCaS generates an AUC score 0.862. The LabCaS program is freely available at: http://www.csbio.sjtu.edu.cn/bioinf/LabCaS . Proteins 2013. © 2012 Wiley Periodicals, Inc.     

A top-down approach to enhance the power of predicting human protein subcellular localization: Hum-mPLoc 2.0

Predicting subcellular localization of human proteins is a challenging problem, particularly when query proteins may have a multiplex character, i.e., simultaneously exist at, or move between, two or more different subcellular location sites. In a previous study, we developed a predictor called “Hum-mPLoc” to deal with the multiplex problem for the human protein system. However, Hum-mPLoc has the following shortcomings. (1) The input of accession number for a query protein is required in order to obtain a higher expected success rate by selecting to use the higher-level prediction pathway; but many proteins, such as synthetic and hypothetical proteins as well as those newly discovered proteins without being deposited into databanks yet, do not have accession numbers. (2) Neither functional domain nor sequential evolution information were taken into account in Hum-mPLoc, and hence its power may be reduced accordingly. In view of this, a top-down strategy to address these shortcomings has been implemented. The new predictor thus obtained is called Hum-mPLoc 2.0, where the accession number for input is no longer needed whatsoever. Moreover, both the functional domain information and the sequential evolution information have been fused into the predictor by an ensemble classifier. As a consequence, the prediction power has been significantly enhanced. The web server of Hum-mPLoc2.0 is freely accessible at http://www.csbio.sjtu.edu.cn/bioinf/hum-multi-2/.     

Gneg-mPLoc: A top-down strategy to enhance the quality of predicting subcellular localization of Gram-negative bacterial proteins

By incorporating the information of gene ontology, functional domain, and sequential evolution, a new predictor called Gneg-mPLoc was developed. It can be used to identify Gram-negative bacterial proteins among the following eight locations: (1) cytoplasm, (2) extracellular, (3) fimbrium, (4) flagellum, (5) inner membrane, (6) nucleoid, (7) outer membrane, and (8) periplasm. It can also be used to deal with the case when a query protein may simultaneously exist in more than one location. Compared with the original predictor called Gneg-PLoc, the new predictor is much more powerful and flexible. For a newly constructed stringent benchmark dataset in which none of proteins included has ≥25% pairwise sequence identity to any other in a same subset (location), the overall jackknife success rate achieved by Gneg-mPLoc was 85.5%, which was more than 14% higher than the corresponding rate by the Gneg-PLoc. As a user friendly web-server, Gneg-mPLoc is freely accessible at http://www.csbio.sjtu.edu.cn/bioinf/Gneg-multi/.     

Cryo-EM image alignment: From pair-wise to joint with deep unsupervised difference learning

Cryo-electron microscopy (cryo-EM) single-particle analysis is a revolutionary imaging technique to resolve and visualize biomacromolecules. Image alignment in cryo-EM is an important and basic step to improve the precision of the image distance calculation. However, it is a very challenging task due to high noise and low signal-to-noise ratio. Therefore, we propose a new deep unsupervised difference learning (UDL) strategy with novel pseudo-label guided learning network architecture and apply it to pair-wise image alignment in cryo-EM. The training framework is fully unsupervised. Furthermore, a variant of UDL called joint UDL (JUDL), is also proposed, which is capable of utilizing the similarity information of the whole dataset and thus further increase the alignment precision. Assessments on both real-world and synthetic cryo-EM single-particle image datasets suggest the new unsupervised joint alignment method can achieve more accurate alignment results. Our method is highly efficient by taking advantages of GPU devices. The source code of our methods is publicly available at “http://www.csbio.sjtu.edu.cn/bioinf/JointUDL/” for academic use.     

Improving accuracy of protein contact prediction using balanced network deconvolution

Residue contact map is essential for protein three‐dimensional structure determination. But most of the current contact prediction methods based on residue co‐evolution suffer from high false‐positives as introduced by indirect and transitive contacts (i.e., residues A–B and B–C are in contact, but A–C are not). Built on the work by Feizi et al. (Nat Biotechnol 2013; 31:726–733), which demonstrated a general network model to distinguish direct dependencies by network deconvolution, this study presents a new balanced network deconvolution (BND) algorithm to identify optimized dependency matrix without limit on the eigenvalue range in the applied network systems. The algorithm was used to filter contact predictions of five widely used co‐evolution methods. On the test of proteins from three benchmark datasets of the 9th critical assessment of protein structure prediction (CASP9), CASP10, and PSICOV (precise structural contact prediction using sparse inverse covariance estimation) database experiments, the BND can improve the medium‐ and long‐range contact predictions at the L/5 cutoff by 55.59% and 47.68%, respectively, without additional central processing unit cost. The improvement is statistically significant, with a P‐value < 5.93 × 10^?3 in the Student's t‐test. A further comparison with the ab initio structure predictions in CASPs showed that the usefulness of the current co‐evolution‐based contact prediction to the three‐dimensional structure modeling relies on the number of homologous sequences existing in the sequence databases. BND can be used as a general contact refinement method, which is freely available at: http://www.csbio.sjtu.edu.cn/bioinf/BND/ . Proteins 2015; 83:485–496. © 2014 Wiley Periodicals, Inc.     

Identification of proteases and their types

Called by many as biology’s version of Swiss army knives, proteases cut long sequences of amino acids into fragments and regulate most physiological processes. They are vitally important in the life cycle. Different types of proteases have different action mechanisms and biological processes. With the avalanche of protein sequences generated during the postgenomic age, it is highly desirable for both basic research and drug design to develop a fast and reliable method for identifying the types of proteases according to their sequences or even just for whether they are proteases or not. In this article, three recently developed identification methods in this regard are discussed: (i) FunD-PseAAC, (ii) GO-PseAAC, and (iii) FunD-PsePSSM. The first two were established by hybridizing the FunD (functional domain) approach and the GO (gene ontology) approach, respectively, with the PseAAC (pseudo amino acid composition) approach. The third method was established by fusing the FunD approach with the PsePSSM (pseudo position-specific scoring matrix) approach. Of these three methods, only FunD-PsePSSM has provided a server called ProtIdent (protease identifier), which is freely accessible to the public via the website at http://www.csbio.sjtu.edu.cn/bioinf/Protease. For the convenience of users, a step-by-step guide on how to use ProtIdent is illustrated. Meanwhile, the caveat in using ProtIdent and how to understand the success expectancy rate of a statistical predictor are discussed. Finally, the essence of why ProtIdent can yield a high success rate in identifying proteases and their types is elucidated.     

LigBind: Identifying Binding Residues for Over 1000 Ligands with Relation-Aware Graph Neural Networks

Identifying the interactions between proteins and ligands is significant for drug discovery and design. Considering the diverse binding patterns of ligands, the ligand-specific methods are trained per ligand to predict binding residues. However, most of the existing ligand-specific methods ignore shared binding preferences among various ligands and generally only cover a limited number of ligands with a sufficient number of known binding proteins. In this study, we propose a relation-aware framework LigBind with graph-level pre-training to enhance the ligand-specific binding residue predictions for 1159 ligands, which can effectively cover the ligands with a few known binding proteins. LigBind first pre-trains a graph neural network-based feature extractor for ligand-residue pairs and relation-aware classifiers for similar ligands. Then, LigBind is fine-tuned with ligand-specific binding data, where a domain adaptive neural network is designed to automatically leverage the diversity and similarity of various ligand-binding patterns for accurate binding residue prediction. We construct ligand-specific benchmark datasets of 1159 ligands and 16 unseen ligands, which are used to evaluate the effectiveness of LigBind. The results demonstrate the LigBind’s efficacy on large-scale ligand-specific benchmark datasets, and it generalizes well to unseen ligands. LigBind also enables accurate identification of the ligand-binding residues in the main protease, papain-like protease and the RNA-dependent RNA polymerase of SARS-CoV-2. The web server and source codes of LigBind are available at http://www.csbio.sjtu.edu.cn/bioinf/LigBind/ and https://github.com/YYingXia/LigBind/ for academic use.     

Towards better accuracy for missing value estimation of epistatic miniarray profiling data by a novel ensemble approach

Epistatic miniarray profiling (E-MAP) is a powerful tool for analyzing gene functions and their biological relevance. However, E-MAP data suffers from large proportion of missing values, which often results in misleading and biased analysis results. It is urgent to develop effective missing value estimation methods for E-MAP. Although several independent algorithms can be applied to achieve this goal, their performance varies significantly on different datasets, indicating different algorithms having their own advantages and disadvantages. In this paper, we propose a novel ensemble approach EMDI based on the high-level diversity to impute missing values that consists of two global and four local base estimators. Experimental results on five E-MAP datasets show that EMDI outperforms all single base algorithms, demonstrating an appropriate combination providing complementarity among different methods. Comparison results between several fusion strategies also demonstrate that the proposed high-level diversity scheme is superior to others. EMDI is freely available at www.csbio.sjtu.edu.cn/bioinf/EMDI/.     

SOMPNN: an efficient non-parametric model for predicting transmembrane helices

Accurately predicting the transmembrane helices (TMH) in a helical membrane protein is an important but challenging task. Recent researches have demonstrated that statistics-based methods are promising routes to improve the TMH prediction accuracy. However, most of existing TMH predictors are parametric models and they have to make assumptions of several or even hundreds of adjustable parameters based on the underlying probability distribution, which is difficult when no a priori knowledge is available. Besides the performances of these parametric predictors significantly depend on the estimated parameters, some of them need to exploit the entire training dataset in the prediction stage, which will lead to low prediction efficiency and this problem will become even worse when dealing with large-scale dataset. In this paper, we propose a novel SOMPNN model for prediction of TMH that features by minimal parameter assumptions requirement and high computational efficiency. In the SOMPNN model, a self-organizing map (SOM) is used to adaptively learn the helices distribution knowledge hidden in the training data, and then a probabilistic neural network (PNN) is adopted to predict TMH segments based on the knowledge learned by SOM. Experimental results on two benchmark datasets show that the proposed SOMPNN outperforms most existing popular TMH predictors and is promising to be extended to deal with other complicated biological problems. The datasets and the source codes of SOMPNN are available at http://www.csbio.sjtu.edu.cn/bioinf/SOMPNN/.     

Support vector machine approach for protein subcellular localization prediction

MOTIVATION: Subcellular localization is a key functional characteristic of proteins. A fully automatic and reliable prediction system for protein subcellular localization is needed, especially for the analysis of large-scale genome sequences. RESULTS: In this paper, Support Vector Machine has been introduced to predict the subcellular localization of proteins from their amino acid compositions. The total prediction accuracies reach 91.4% for three subcellular locations in prokaryotic organisms and 79.4% for four locations in eukaryotic organisms. Predictions by our approach are robust to errors in the protein N-terminal sequences. This new approach provides superior prediction performance compared with existing algorithms based on amino acid composition and can be a complementary method to other existing methods based on sorting signals. AVAILABILITY: A web server implementing the prediction method is available at http://www.bioinfo.tsinghua.edu.cn/SubLoc/. SUPPLEMENTARY INFORMATION: Supplementary material is available at http://www.bioinfo.tsinghua.edu.cn/SubLoc/.