PPRODO: prediction of protein domain boundaries using neural networks

Successful prediction of protein domain boundaries provides valuable information not only for the computational structure prediction of multidomain proteins but also for the experimental structure determination. Since protein sequences of multiple domains may contain much information regarding evolutionary processes such as gene-exon shuffling, this information can be detected by analyzing the position-specific scoring matrix (PSSM) generated by PSI-BLAST. We have presented a method, PPRODO (Prediction of PROtein DOmain boundaries) that predicts domain boundaries of proteins from sequence information by a neural network. The network is trained and tested using the values obtained from the PSSM generated by PSI-BLAST. A 10-fold cross-validation technique is performed to obtain the parameters of neural networks using a nonredundant set of 522 proteins containing 2 contiguous domains. PPRODO provides good and consistent results for the prediction of domain boundaries, with accuracy of about 66% using the +/-20 residue criterion. The PPRODO source code, as well as all data sets used in this work, are available from http://gene.kias.re.kr/ approximately jlee/pprodo/.     

基于位置特异性谱和输入加权神经网络的蛋白质亚细胞定位预测

蛋白质必须处于正确的亚细胞位置才能行使其功能。文章利用PSI-BLAST工具搜索蛋白质序列,提取位点特异性谱中的位点特异性得分矩阵作为蛋白质的一类特征,并计算4等分序列的氨基酸含量以及1～7阶二肽含量作为另外两类特征,由这三类特征一共得到蛋白质序列的12个特征向量。通过设计一个简单加权函数对各类特征向量加权处理,作为神经网络预测器的输入,并使用Levenberg-Marquardt算法代替传统的EBP算法来调整网络权值和阈值,大大提高了训练速度。对具有4类亚细胞位置和12类亚细胞位置的两种蛋白质数据集分别进行＂留一法＂测试和5倍交叉验证测试,总体预测精度分别达到88.4%和83.3%。其中,对4类亚细胞位置数据集的预测效果优于普通BP神经网络、隐马尔可夫模型、模糊K邻近等预测方法,对12类亚细胞位置数据集的预测效果优于支持向量机分类方法。最后还对三类特征采取不同加权比例对预测精度的影响进行了讨论,对选择的八种加权比例的预测结果表明,分别给予三类特征合适的权值系数可以进一步提高预测精度。     

A graphic representation of protein sequence and predicting the subcellular locations of prokaryotic proteins

Zp curve, a three-dimensional space curve representation of protein primary sequence based on the hydrophobicity and charged properties of amino acid residues along the primary sequence is suggested. Relying on the Zp parameters extracted from the three components of the Zp curve and the Bayes discriminant algorithm, the subcellular locations of prokaryotic proteins were predicted. Consequently, an accuracy of 81.5% in the cross-validation test has been achieved using 13 parameters extracted from the curve for the database of 997 prokaryotic proteins. The result is slightly better than that of using the neural network method (80.9%) based on the amino acid composition for the same database. By jointing the amino acid composition and the Zp parameters, the overall predictive accuracy 89.6% can be achieved. It is about 3% higher than that of the Bayes discriminant algorithm based merely on the amino acid composition for the same database. The prediction is also performed with a larger dataset derived from the version 39 SWISS-PROT databank and two datasets with different sequence similarity. Even for the dataset of non-sequence similarity, the improvement can be of 4.4% in the cross-validation test. The results indicate that the Zp parameters are effective in representing the information within a protein primary sequence. The method of extracting information from the primary structure may be useful for other areas of protein studies.     

SPIN2: Predicting sequence profiles from protein structures using deep neural networks

Designing protein sequences that can fold into a given structure is a well‐known inverse protein‐folding problem. One important characteristic to attain for a protein design program is the ability to recover wild‐type sequences given their native backbone structures. The highest average sequence identity accuracy achieved by current protein‐design programs in this problem is around 30%, achieved by our previous system, SPIN. SPIN is a program that predicts sequences compatible with a provided structure using a neural network with fragment‐based local and energy‐based nonlocal profiles. Our new model, SPIN2, uses a deep neural network and additional structural features to improve on SPIN. SPIN2 achieves over 34% in sequence recovery in 10‐fold cross‐validation and independent tests, a 4% improvement over the previous version. The sequence profiles generated from SPIN2 are expected to be useful for improving existing fold recognition and protein design techniques. SPIN2 is available at http://sparks-lab.org .     

Domain position prediction based on sequence information by using fuzzy mean operator

The prediction of protein domain region is an advantageous process on the study of protein structure and function. In this study, we proposed a new method, which is composed of fuzzy mean operator and region division, to predict the particular positions of domains in a target protein based on its sequence. The whole sequence is aligned and scored by using fuzzy mean operator, and the final determination of domain region position is realized by region division. A published benchmark is used for the comparison with previous researches. In addition, we generate two extra datasets to examine the stability of this method. Finally, the prediction accuracy of independent test dataset achieved by our method was up to 84.13%. We wish that this method could be useful for related researches. Proteins 2015; 83:1462–1469. © 2015 Wiley Periodicals, Inc.     

An improved profile-level domain linker propensity index for protein domain boundary prediction

Protein domain boundary prediction is critical for understanding protein structure and function. In this study, we present a novel method, an order profile domain linker propensity index (OPI), which uses the evolutionary information extracted from the protein sequence frequency profiles calculated from the multiple sequence alignments. A protein sequence is first converted into smooth and normalized numeric order profiles by OPI, from which the domain linkers can be predicted. By discriminating the different frequencies of the amino acids in the protein sequence frequency profiles, OPI clearly shows better performance than our previous method, a binary profile domain linker propensity index (PDLI). We tested our new method on two different datasets, SCOP-1 dataset and SCOP-2 dataset, and we were able to achieve a precision of 0.82 and 0.91 respectively. OPI also outperforms other residue-level, profile-level indexes as well as other state-of-the-art methods.     

A combination of compositional index and genetic algorithm for predicting transmembrane helical segments

Transmembrane helix (TMH) topology prediction is becoming a focal problem in bioinformatics because the structure of TM proteins is difficult to determine using experimental methods. Therefore, methods that can computationally predict the topology of helical membrane proteins are highly desirable. In this paper we introduce TMHindex, a method for detecting TMH segments using only the amino acid sequence information. Each amino acid in a protein sequence is represented by a Compositional Index, which is deduced from a combination of the difference in amino acid occurrences in TMH and non-TMH segments in training protein sequences and the amino acid composition information. Furthermore, a genetic algorithm was employed to find the optimal threshold value for the separation of TMH segments from non-TMH segments. The method successfully predicted 376 out of the 378 TMH segments in a dataset consisting of 70 test protein sequences. The sensitivity and specificity for classifying each amino acid in every protein sequence in the dataset was 0.901 and 0.865, respectively. To assess the generality of TMHindex, we also tested the approach on another standard 73-protein 3D helix dataset. TMHindex correctly predicted 91.8% of proteins based on TM segments. The level of the accuracy achieved using TMHindex in comparison to other recent approaches for predicting the topology of TM proteins is a strong argument in favor of our proposed method. Availability: The datasets, software together with supplementary materials are available at: http://faculty.uaeu.ac.ae/nzaki/TMHindex.htm.     

Inferring RNA-binding protein target preferences using adversarial domain adaptation

Precise identification of target sites of RNA-binding proteins (RBP) is important to understand their biochemical and cellular functions. A large amount of experimental data is generated by in vivo and in vitro approaches. The binding preferences determined from these platforms share similar patterns but there are discernable differences between these datasets. Computational methods trained on one dataset do not always work well on another dataset. To address this problem which resembles the classic “domain shift” in deep learning, we adopted the adversarial domain adaptation (ADDA) technique and developed a framework (RBP-ADDA) that can extract RBP binding preferences from an integration of in vivo and vitro datasets. Compared with conventional methods, ADDA has the advantage of working with two input datasets, as it trains the initial neural network for each dataset individually, projects the two datasets onto a feature space, and uses an adversarial framework to derive an optimal network that achieves an optimal discriminative predictive power. In the first step, for each RBP, we include only the in vitro data to pre-train a source network and a task predictor. Next, for the same RBP, we initiate the target network by using the source network and use adversarial domain adaptation to update the target network using both in vitro and in vivo data. These two steps help leverage the in vitro data to improve the prediction on in vivo data, which is typically challenging with a lower signal-to-noise ratio. Finally, to further take the advantage of the fused source and target data, we fine-tune the task predictor using both data. We showed that RBP-ADDA achieved better performance in modeling in vivo RBP binding data than other existing methods as judged by Pearson correlations. It also improved predictive performance on in vitro datasets. We further applied augmentation operations on RBPs with less in vivo data to expand the input data and showed that it can improve prediction performances. Lastly, we explored the predictive interpretability of RBP-ADDA, where we quantified the contribution of the input features by Integrated Gradients and identified nucleotide positions that are important for RBP recognition.     

Predicting protein solubility with a hybrid approach by pseudo amino acid composition

Protein solubility plays a major role for understanding the crystal growth and crystallization process of protein. How to predict the propensity of a protein to be soluble or to form inclusion body is a long but not fairly resolved problem. After choosing almost 10,000 protein sequences from NCBI database and eliminating the sequences with 90% homologous similarity by CD-HIT, 5692 sequences remained. By using Chou's pseudo amino acid composition features, we predict the soluble protein with the three methods: support vector machine (SVM), back propagation neural network (BP Neural Network) and hybrid method based on SVM and BP Neural Network, respectively. Each method is evaluated by re-substitution test and 10-fold cross-validation test. In the re-substitution test, the BP Neural Network performs with the best results, in which the accuracy achieves 0.9288 and Matthews Correlation Coefficient (MCC) achieves 0.8513. Meanwhile, the other two methods are better than BP Neural Network in 10-fold cross-validation test. The hybrid method based on SVM and BP Neural Network is the best. The average accuracy is 0.8678 and average MCC is 0.7233. Although all of the three methods achieve considerable evaluations, the hybrid method is deemed to be the best, according to the performance comparison.     

Protein subcellular localization based on PSI-BLAST and machine learning

Subcellular location is an important functional annotation of proteins. An automatic, reliable and efficient prediction system for protein subcellular localization is necessary for large-scale genome analysis. This paper describes a protein subcellular localization method which extracts features from protein profiles rather than from amino acid sequences. The protein profile represents a protein family, discards part of the sequence information that is not conserved throughout the family and therefore is more sensitive than the amino acid sequence. The amino acid compositions of whole profile and the N-terminus of the profile are extracted, respectively, to train and test the probabilistic neural network classifiers. On two benchmark datasets, the overall accuracies of the proposed method reach 89.1% and 68.9%, respectively. The prediction results show that the proposed method perform better than those methods based on amino acid sequences. The prediction results of the proposed method are also compared with Subloc on two redundance-reduced datasets.     

3dswap-pred: prediction of 3D domain swapping from protein sequence using Random Forest approach

3D domain swapping is a protein structural phenomenon that mediates the formation of the higher order oligomers in a variety of proteins with different structural and functional properties. 3D domain swapping is associated with a variety of biological functions ranging from oligomerization to pathological conformational diseases. 3D domain swapping is realised subsequent to structure determination where the protein is observed in the swapped conformation in the oligomeric state. This is a limiting step to understand this important structural phenomenon in a large scale from the growing sequence data. A new machine learning approach, 3dswap-pred, has been developed for the prediction of 3D domain swapping in protein structures from mere sequence data using the Random Forest approach. 3Dswap-pred is implemented using a positive sequence dataset derived from literature based structural curation of 297 structures. A negative sequence dataset is obtained from 462 SCOP domains using a new sequence data mining approach and a set of 126 sequencederived features. Statistical validation using an independent dataset of 68 positive sequences and 313 negative sequences revealed that 3dswap-pred achieved an accuracy of 63.8%. A webserver is also implemented using the 3dswap-pred Random Forest model. The server is available from the URL: http://caps.ncbs.res.in/3dswap-pred.     

Clustering of multi‐domain protein sequences

The overall function of a multi‐domain protein is determined by the functional and structural interplay of its constituent domains. Traditional sequence alignment‐based methods commonly utilize domain‐level information and provide classification only at the level of domains. Such methods are not capable of taking into account the contributions of other domains in the proteins, and domain‐linker regions and classify multi‐domain proteins. An alignment‐free protein sequence comparison tool, CLAP (CLAssification of Proteins) was previously developed in our laboratory to especially handle multi‐domain protein sequences without a requirement of defining domain boundaries and sequential order of domains. Through this method we aim to achieve a biologically meaningful classification scheme for multi‐domain protein sequences. In this article, CLAP‐based classification has been explored on 5 datasets of multi‐domain proteins and we present detailed analysis for proteins containing (1) Tyrosine phosphatase and (2) SH3 domain. At the domain‐level CLAP‐based classification scheme resulted in a clustering similar to that obtained from an alignment‐based method. CLAP‐based clusters obtained for full‐length datasets were shown to comprise of proteins with similar functions and domain architectures. Our study demonstrates that multi‐domain proteins could be classified effectively by considering full‐length sequences without a requirement of identification of domains in the sequence.     

Predicting subcellular localization of proteins by hybridizing functional domain composition and pseudo-amino acid composition

Recent advances in large-scale genome sequencing have led to the rapid accumulation of amino acid sequences of proteins whose functions are unknown. Since the functions of these proteins are closely correlated with their subcellular localizations, many efforts have been made to develop a variety of methods for predicting protein subcellular location. In this study, based on the strategy by hybridizing the functional domain composition and the pseudo-amino acid composition (Cai and Chou [2003]: Biochem. Biophys. Res. Commun. 305:407-411), the Intimate Sorting Algorithm (ISort predictor) was developed for predicting the protein subcellular location. As a showcase, the same plant and non-plant protein datasets as investigated by the previous investigators were used for demonstration. The overall success rate by the jackknife test for the plant protein dataset was 85.4%, and that for the non-plant protein dataset 91.9%. These are so far the highest success rates achieved for the two datasets by following a rigorous cross validation test procedure, further confirming that such a hybrid approach may become a very useful high-throughput tool in the area of bioinformatics, proteomics, as well as molecular cell biology.     

Prediction of MHC binding peptide using Gibbs motif sampler,weight matrix and artificial neural network

The identification of MHC restricted epitopes is an important goal in peptide based vaccine and diagnostic development. As wet lab experiments for identification of MHC binding peptide are expensive and time consuming, in silico tools have been developed as fast alternatives, however with low performance. In the present study, we used IEDB training and blind validation datasets for the prediction of peptide binding to fourteen human MHC class I and II molecules using Gibbs motif sampler, weight matrix and artificial neural network methods. As compare to MHC class I predictor based on sequence weighting (Aroc=0.95 and CC=0.56) and artificial neural network (Aroc=0.73 and CC=0.25), MHC class II predictor based on Gibbs sampler did not perform well (Aroc=0.62 and CC=0.19). The predictive accuracy of Gibbs motif sampler in identifying the 9-mer cores of a binding peptide to DRB1 alleles are also limited (40¢), however above the random prediction (14¢). Therefore, the size of dataset (training and validation) and the correct identification of the binding core are the two main factors limiting the performance of MHC class-II binding peptide prediction. Overall, these data suggest that there is substantial room to improve the quality of the core predictions using novel approaches that capture distinct features of MHC-peptide interactions than the current approaches.     

iDHS-DSAMS: Identifying DNase I hypersensitive sites based on the dinucleotide property matrix and ensemble bagged tree

DNase I hypersensitive site (DHS) is related to DNA regulatory elements, so the understanding of DHS sites is of great significance for biomedical research. However, traditional experiments are not very good at identifying recombinant sites of a large number of emerging DNA sequences by sequencing. Some machine learning methods have been proposed to identify DHS, but most methods ignore spatial autocorrelation of the DNA sequence. In this paper, we proposed a predictor called iDHS-DSAMS to identify DHS based on the benchmark datasets. We develop a feature extraction method called dinucleotide-based spatial autocorrelation (DSA). Then we use Min-Redundancy-Max-Relevance (mRMR) to remove irrelevant and redundant features and a 100-dimensional feature vector is selected. Finally, we utilize ensemble bagged tree as classifier, which is based on the oversampled datasets using SMOTE. Five-fold cross validation tests on two benchmark datasets indicate that the proposed method outperforms its existing counterparts on the individual accuracy (Acc), Matthews correlation coefficient (MCC), sensitivity (Sn) and specificity (Sp).     

Classifier ensembles for protein structural class prediction with varying homology

Structural class characterizes the overall folding type of a protein or its domain. A number of computational methods have been proposed to predict structural class based on primary sequences; however, the accuracy of these methods is strongly affected by sequence homology. This paper proposes, an ensemble classification method and a compact feature-based sequence representation. This method improves prediction accuracy for the four main structural classes compared to competing methods, and provides highly accurate predictions for sequences of widely varying homologies. The experimental evaluation of the proposed method shows superior results across sequences that are characterized by entire homology spectrum, ranging from 25% to 90% homology. The error rates were reduced by over 20% when compared with using individual prediction methods and most commonly used composition vector representation of protein sequences. Comparisons with competing methods on three large benchmark datasets consistently show the superiority of the proposed method.     

Computational prediction of conformational B-cell epitopes from antigen primary structures by ensemble learning

Motivation

The conformational B-cell epitopes are the specific sites on the antigens that have immune functions. The identification of conformational B-cell epitopes is of great importance to immunologists for facilitating the design of peptide-based vaccines. As an attempt to narrow the search for experimental validation, various computational models have been developed for the epitope prediction by using antigen structures. However, the application of these models is undermined by the limited number of available antigen structures. In contrast to the most of available structure-based methods, we here attempt to accurately predict conformational B-cell epitopes from antigen sequences.

Methods

In this paper, we explore various sequence-derived features, which have been observed to be associated with the location of epitopes or ever used in the similar tasks. These features are evaluated and ranked by their discriminative performance on the benchmark datasets. From the perspective of information science, the combination of various features can usually lead to better results than the individual features. In order to build the robust model, we adopt the ensemble learning approach to incorporate various features, and develop the ensemble model to predict conformational epitopes from antigen sequences.

Results

Evaluated by the leave-one-out cross validation, the proposed method gives out the mean AUC scores of 0.687 and 0.651 on two datasets respectively compiled from the bound structures and unbound structures. When compared with publicly available servers by using the independent dataset, our method yields better or comparable performance. The results demonstrate the proposed method is useful for the sequence-based conformational epitope prediction.

Availability

The web server and datasets are freely available at http://bcell.whu.edu.cn.     

Armadillo: domain boundary prediction by amino acid composition

The identification and annotation of protein domains provides a critical step in the accurate determination of molecular function. Both computational and experimental methods of protein structure determination may be deterred by large multi-domain proteins or flexible linker regions. Knowledge of domains and their boundaries may reduce the experimental cost of protein structure determination by allowing researchers to work on a set of smaller and possibly more successful alternatives. Current domain prediction methods often rely on sequence similarity to conserved domains and as such are poorly suited to detect domain structure in poorly conserved or orphan proteins. We present here a simple computational method to identify protein domain linkers and their boundaries from sequence information alone. Our domain predictor, Armadillo (http://armadillo.blueprint.org), uses any amino acid index to convert a protein sequence to a smoothed numeric profile from which domains and domain boundaries may be predicted. We derived an amino acid index called the domain linker propensity index (DLI) from the amino acid composition of domain linkers using a non-redundant structure dataset. The index indicates that Pro and Gly show a propensity for linker residues while small hydrophobic residues do not. Armadillo predicts domain linker boundaries from Z-score distributions and obtains 35% sensitivity with DLI in a two-domain, single-linker dataset (within +/-20 residues from linker). The combination of DLI and an entropy-based amino acid index increases the overall Armadillo sensitivity to 56% for two domain proteins. Moreover, Armadillo achieves 37% sensitivity for multi-domain proteins, surpassing most other prediction methods. Armadillo provides a simple, but effective method by which prediction of domain boundaries can be obtained with reasonable sensitivity. Armadillo should prove to be a valuable tool for rapidly delineating protein domains in poorly conserved proteins or those with no sequence neighbors. As a first-line predictor, domain meta-predictors could yield improved results with Armadillo predictions.     

Integrated gene and species phylogenies from unaligned whole genome protein sequences

MOTIVATION: Most molecular phylogenies are based on sequence alignments. Consequently, they fail to account for modes of sequence evolution that involve frequent insertions or deletions. Here we present a method for generating accurate gene and species phylogenies from whole genome sequence that makes use of short character string matches not placed within explicit alignments. In this work, the singular value decomposition of a sparse tetrapeptide frequency matrix is used to represent the proteins of organisms uniquely and precisely as vectors in a high-dimensional space. Vectors of this kind can be used to calculate pairwise distance values based on the angle separating the vectors, and the resulting distance values can be used to generate phylogenetic trees. Protein trees so derived can be examined directly for homologous sequences. Alternatively, vectors defining each of the proteins within an organism can be summed to provide a vector representation of the organism, which is then used to generate species trees. RESULTS: Using a large mitochondrial genome dataset, we have produced species trees that are largely in agreement with previously published trees based on the analysis of identical datasets using different methods. These trees also agree well with currently accepted phylogenetic theory. In principle, our method could be used to compare much larger bacterial or nuclear genomes in full molecular detail, ultimately allowing accurate gene and species relationships to be derived from a comprehensive comparison of complete genomes. In contrast to phylogenetic methods based on alignments, sequences that evolve by relative insertion or deletion would tend to remain recognizably similar.     

A novel neural response algorithm for protein function prediction

Background

Large amounts of data are being generated by high-throughput genome sequencing methods. But the rate of the experimental functional characterization falls far behind. To fill the gap between the number of sequences and their annotations, fast and accurate automated annotation methods are required. Many methods, such as GOblet, GOFigure, and Gotcha, are designed based on the BLAST search. Unfortunately, the sequence coverage of these methods is low as they cannot detect the remote homologues. Adding to this, the lack of annotation specificity advocates the need to improve automated protein function prediction.

Results

We designed a novel automated protein functional assignment method based on the neural response algorithm, which simulates the neuronal behavior of the visual cortex in the human brain. Firstly, we predict the most similar target protein for a given query protein and thereby assign its GO term to the query sequence. When assessed on test set, our method ranked the actual leaf GO term among the top 5 probable GO terms with accuracy of 86.93%.

Conclusions

The proposed algorithm is the first instance of neural response algorithm being used in the biological domain. The use of HMM profiles along with the secondary structure information to define the neural response gives our method an edge over other available methods on annotation accuracy. Results of the 5-fold cross validation and the comparison with PFP and FFPred servers indicate the prominent performance by our method. The program, the dataset, and help files are available at http://www.jjwanglab.org/NRProF/.