Statistical geometry based prediction of nonsynonymous SNP functional effects using random forest and neuro-fuzzy classifiers

There is substantial interest in methods designed to predict the effect of nonsynonymous single nucleotide polymorphisms (nsSNPs) on protein function, given their potential relationship to heritable diseases. Current state-of-the-art supervised machine learning algorithms, such as random forest (RF), train models that classify single amino acid mutations in proteins as either neutral or deleterious to function. However, it is frequently the case that the functional effect of a polymorphism on a protein resides between these two extremes. The utilization of classifiers that incorporate fuzzy logic provides a natural extension in order to account for the spectrum of possible functional consequences. We generated a dataset of single amino acid substitutions in human proteins having known three-dimensional structures. Each variant was uniquely represented as a feature vector that included computational geometry and knowledge-based statistical potential predictors obtained though application of Delaunay tessellation of protein structures. Additional attributes consisted of physicochemical properties of the native and replacement amino acids as well as topological location of the mutated residue position in the solved structure. Classification performance of the RF algorithm was evaluated on a training set consisting of the disease-associated and neutral nsSNPs taken from our dataset, and attributes were ranked according to their relative importance. Similarly, we evaluated the performance of adaptive neuro-fuzzy inference system (ANFIS). The utility of statistical geometry predictors was compared with that of traditional structural and evolutionary attributes employed by other researchers, revealing an equally effective yet complementary methodology. Among all attributes in our feature set, the statistical geometry predictors were found to be the most highly ranked. On the basis of the AUC (area under the ROC curve) measure of performance, the ANFIS and RF models were equally effective when only statistical geometry features were utilized. Tenfold cross-validation studies evaluating AUC, balanced error rate (BER), and Matthew's correlation coefficient (MCC) showed that our RF model was at least comparable with the well-established methods of SIFT and PolyPhen. The trained RF and ANFIS models were each subsequently used to predict the disease potential of human nsSNPs in our dataset that are currently unclassified (http://rna.gmu.edu/FuzzySnps/).     

Evaluation of structural and evolutionary contributions to deleterious mutation prediction

Methods for automated prediction of deleterious protein mutations have utilized both structural and evolutionary information but the relative contribution of these two factors remains unclear. To address this, we have used a variety of structural and evolutionary features to create simple deleterious mutation models that have been tested on both experimental mutagenesis and human allele data. We find that the most accurate predictions are obtained using a solvent-accessibility term, the C(beta) density, and a score derived from homologous sequences, SIFT. A classification tree using these two features has a cross-validated prediction error of 20.5% on an experimental mutagenesis test set when the prior probability for deleterious and neutral cases is equal, whereas this prediction error is 28.8% and 22.2% using either the C(beta) density or SIFT alone. The improvement imparted by structure increases when fewer homologs are available: when restricted to three homologs the prediction error improves from 26.9% using SIFT alone to 22.4% using SIFT and the C(beta) density, or 24.8% using SIFT and a noisy C(beta) density term approximating the inaccuracy of ab initio structures modeled by the Rosetta method. We conclude that methods for deleterious mutation prediction should include structural information when fewer than five to ten homologs are available, and that ab initio predicted structures may soon be useful in such cases when high-resolution structures are unavailable.     

Predicting the insurgence of human genetic diseases associated to single point protein mutations with support vector machines and evolutionary information

MOTIVATION: Human single nucleotide polymorphisms (SNPs) are the most frequent type of genetic variation in human population. One of the most important goals of SNP projects is to understand which human genotype variations are related to Mendelian and complex diseases. Great interest is focused on non-synonymous coding SNPs (nsSNPs) that are responsible of protein single point mutation. nsSNPs can be neutral or disease associated. It is known that the mutation of only one residue in a protein sequence can be related to a number of pathological conditions of dramatic social impact such as Alzheimer's, Parkinson's and Creutzfeldt-Jakob's diseases. The quality and completeness of presently available SNPs databases allows the application of machine learning techniques to predict the insurgence of human diseases due to single point protein mutation starting from the protein sequence. RESULTS: In this paper, we develop a method based on support vector machines (SVMs) that starting from the protein sequence information can predict whether a new phenotype derived from a nsSNP can be related to a genetic disease in humans. Using a dataset of 21 185 single point mutations, 61% of which are disease-related, out of 3587 proteins, we show that our predictor can reach more than 74% accuracy in the specific task of predicting whether a single point mutation can be disease related or not. Our method, although based on less information, outperforms other web-available predictors implementing different approaches. AVAILABILITY: A beta version of the web tool is available at http://gpcr.biocomp.unibo.it/cgi/predictors/PhD-SNP/PhD-SNP.cgi     

Prediction by graph theoretic measures of structural effects in proteins arising from non-synonymous single nucleotide polymorphisms

Recent analyses of human genome sequences have given rise to impressive advances in identifying non-synonymous single nucleotide polymorphisms (nsSNPs). By contrast, the annotation of nsSNPs and their links to diseases are progressing at a much slower pace. Many of the current approaches to analysing disease-associated nsSNPs use primarily sequence and evolutionary information, while structural information is relatively less exploited. In order to explore the potential of such information, we developed a structure-based approach, Bongo (Bonds ON Graph), to predict structural effects of nsSNPs. Bongo considers protein structures as residue-residue interaction networks and applies graph theoretical measures to identify the residues that are critical for maintaining structural stability by assessing the consequences on the interaction network of single point mutations. Our results show that Bongo is able to identify mutations that cause both local and global structural effects, with a remarkably low false positive rate. Application of the Bongo method to the prediction of 506 disease-associated nsSNPs resulted in a performance (positive predictive value, PPV, 78.5%) similar to that of PolyPhen (PPV, 77.2%) and PANTHER (PPV, 72.2%). As the Bongo method is solely structure-based, our results indicate that the structural changes resulting from nsSNPs are closely associated to their pathological consequences.     

EFIN: predicting the functional impact of nonsynonymous single nucleotide polymorphisms in human genome

Background

Predicting the functional impact of amino acid substitutions (AAS) caused by nonsynonymous single nucleotide polymorphisms (nsSNPs) is becoming increasingly important as more and more novel variants are being discovered. Bioinformatics analysis is essential to predict potentially causal or contributing AAS to human diseases for further analysis, as for each genome, thousands of rare or private AAS exist and only a very small number of which are related to an underlying disease. Existing algorithms in this field still have high false prediction rate and novel development is needed to take full advantage of vast amount of genomic data.

Results

Here we report a novel algorithm that features two innovative changes: 1. making better use of sequence conservation information by grouping the homologous protein sequences into six blocks according to evolutionary distances to human and evaluating sequence conservation in each block independently, and 2. including as many such homologous sequences as possible in analyses. Random forests are used to evaluate sequence conservation in each block and to predict potential impact of an AAS on protein function. Testing of this algorithm on a comprehensive dataset showed significant improvement on prediction accuracy upon currently widely-used programs. The algorithm and a web-based application tool implementing it, EFIN (Evaluation of Functional Impact of Nonsynonymous SNPs) were made freely available (http://paed.hku.hk/efin/) to the public.

Conclusions

Grouping homologous sequences into different blocks according to the evolutionary distance of the species to human and evaluating sequence conservation in each group independently significantly improved prediction accuracy. This approach may help us better understand the roles of genetic variants in human disease and health.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-455) contains supplementary material, which is available to authorized users.     

Prediction of Disease Causing Non-Synonymous SNPs by the Artificial Neural Network Predictor NetDiseaseSNP

We have developed a sequence conservation-based artificial neural network predictor called NetDiseaseSNP which classifies nsSNPs as disease-causing or neutral. Our method uses the excellent alignment generation algorithm of SIFT to identify related sequences and a combination of 31 features assessing sequence conservation and the predicted surface accessibility to produce a single score which can be used to rank nsSNPs based on their potential to cause disease. NetDiseaseSNP classifies successfully disease-causing and neutral mutations. In addition, we show that NetDiseaseSNP discriminates cancer driver and passenger mutations satisfactorily. Our method outperforms other state-of-the-art methods on several disease/neutral datasets as well as on cancer driver/passenger mutation datasets and can thus be used to pinpoint and prioritize plausible disease candidates among nsSNPs for further investigation. NetDiseaseSNP is publicly available as an online tool as well as a web service: http://www.cbs.dtu.dk/services/NetDiseaseSNP     

Combination use of protein–protein interaction network topological features improves the predictive scores of deleterious non-synonymous single-nucleotide polymorphisms

Single-nucleotide polymorphisms (SNPs) are the most frequent form of genetic variations. Non-synonymous SNPs (nsSNPs) occurring in coding region result in single amino acid substitutions that associate with human hereditary diseases. Plenty of approaches were designed for distinguishing deleterious from neutral nsSNPs based on sequence level information. Novel in this work, combinations of protein–protein interaction (PPI) network topological features were introduced in predicting disease-related nsSNPs. Based on a dataset that was compiled from Swiss-Prot, a random forest model was constructed with an average accuracy value of 80.43 % and an MCC value of 0.60 in a rigorous tenfold crossvalidation test. For an independent dataset, our model achieved an accuracy of 88.05 % and an MCC of 0.67. Compared with previous studies, our approach presented superior prediction ability. Results showed that the incorporated PPI network topological features outperform conventional features. Our further analysis indicated that disease-related proteins are topologically different from other proteins. This study suggested that nsSNPs may share some topological information of proteins and the change of topological attributes could provide clues in illustrating functional shift due to nsSNPs.     

Candidate nsSNPs that can affect the functions and interactions of cell cycle proteins

Nonsynonymous single nucleotide polymorphisms (nsSNPs) alter the encoded amino acid sequence, and are thus likely to affect the function of the proteins, and represent potential disease-modifiers. There is an enormous number of nsSNPs in the human population, and the major challenge lies in distinguishing the functionally significant and potentially disease-related ones from the rest. In this study, we analyzed the genetic variations that can alter the functions and the interactions of a group of cell cycle proteins (n = 60) and the proteins interacting with them (n = 26) using computational tools. As a result, we extracted 249 nsSNPs from 77 cell cycle proteins and their interaction partners from public SNP databases. Only 31 (12.4%) of the nsSNPs were validated. The majority (64.5%) of the validated SNPs were rare (minor allele frequencies < 5%). Evolutionary conservation analysis using the SIFT tool suggested that 16.1% of the validated nsSNPs may disrupt the protein function. In addition, 58% of the validated nsSNPs were located in functional protein domains/motifs, which together with the evolutionary conservation analysis enabled us to infer possible biological consequences of the nsSNPs in our set. Our study strongly suggests the presence of naturally occurring genetic variations in the cell cycle proteins that may affect their interactions and functions with possible roles in complex human diseases, such as cancer.     

SMpred: a support vector machine approach to identify structural motifs in protein structure without using evolutionary information

Knowledge of three dimensional structure is essential to understand the function of a protein. Although the overall fold is made from the whole details of its sequence, a small group of residues, often called as structural motifs, play a crucial role in determining the protein fold and its stability. Identification of such structural motifs requires sufficient number of sequence and structural homologs to define conservation and evolutionary information. Unfortunately, there are many structures in the protein structure databases have no homologous structures or sequences. In this work, we report an SVM method, SMpred, to identify structural motifs from single protein structure without using sequence and structural homologs. SMpred method was trained and tested using 132 proteins domains containing 581 motifs. SMpred method achieved 78.79% accuracy with 79.06% sensitivity and 78.53% specificity. The performance of SMpred was evaluated with MegaMotifBase using 188 proteins containing 1161 motifs. Out of 1161 motifs, SMpred correctly identified 1503 structural motifs reported in MegaMotifBase. Further, we showed that SMpred is useful approach for the length deviant superfamilies and single member superfamilies. This result suggests the usefulness of our approach for facilitating the identification of structural motifs in protein structure in the absence of sequence and structural homologs. The dataset and executable for the SMpred algorithm is available at http://www3.ntu.edu.sg/home/EPNSugan/index_files/SMpred.htm.     

Performance of an iterated T-HMM for homology detection

MOTIVATION: Much information about new protein sequences is derived from identifying homologous proteins. Such tasks are difficult when the evolutionary relationships are distant. Some modern methods achieve better results by building a model of a set of related sequences, and then identifying new proteins that fit the model. A further advance was the development of iterative methods that refine the model as more homologs are discovered. These methods are generally limited by ad hoc methods of sequence weighting, neglect of underlying evolutionary relationships and the representation of the set with a single one-size-fits-all model. These limitations are avoided through the use of a Tree hidden Markov model (T-HMM) approach. Our previous work described how a non-iterative version of the T-HMM method could identify distant homologs with superior performance compared with other non-iterated approaches, and described how this method was particularly appropriate for being implemented as an iterative algorithm. RESULTS: We describe an iterative version of the T-HMM algorithm, and evaluate its performance for the detection of distant homologs. Significant improvement over other commonly used methods is found. AVAILABILITY: The software (C++, Perl) is available from the corresponding author.     

Exploration of deleterious single nucleotide polymorphisms in late-onset Alzheimer disease susceptibility genes

Non-synonymous single nucleotide polymorphisms (nsSNPs) are considered as biomarkers to disease susceptibility. In the present study, nsSNPs in CLU, PICALM and BIN1 genes were screened for their functional impact on concerned proteins and their plausible role in Alzheimer disease (AD) susceptibility. Initially, SNPs were retrieved from dbSNP database, followed by identification of potentially deleterious nsSNPs and prediction of their effect on proteins by PolyPhen and SIFT. Protein stability and the probability of mutation occurrence were predicted using I-Mutant and PANTHER respectively. SNPs3D and FASTSNP were used for the functional analysis of nsSNPs. The functional impact on the 3D structure of proteins was evaluated by SWISSPDB viewer and NOMAD-Ref server. On analysis, 3 nsSNPs with IDs rs12800974 (T158P) of PICALM and rs11554585 (R397C) and rs11554585 (N106D) of BIN1 were predicted to be functionally significant with higher scores of I-Mutant, SIFT, PolyPhen, PANTHER, FASTSNP and SNPs3D. The mutant models of these nsSNPs also showed very high energies and RMSD values compared to their native structures. Current study proposes that the three nsSNPs identified in this study constitute a unique resource of potential genetic factors for AD susceptibility.     

An ANN model for the identification of deleterious nsSNPs in tumor suppressor genes

Human genetic variations primarily result from single nucleotide polymorphisms (SNPs) that occurs approximately every 1000 bases in the overall human population. The non-synonymous SNPs (nsSNPs), lead to amino acid changes in the protein product may account for nearly half of the known genetic variations linked to inherited human diseases and cancer. One of the main problems of medical genetics today is to identify nsSNPs that underlie disease-related phenotypes in humans. An attempt was made to develop a new approach to predict such nsSNPs. This would enhance our understanding of genetic diseases and helps to predict the disease. We detect nsSNPs and all possible and reliable alleles by ANN, a soft computing model using potential SNP information. Reliable nsSNPs are identified, based on the reconstructed alleles and on sequence redundancy. The model gives good results with mean specificity (95.85&), sensitivity (97.40&) and accuracy (96.25&). Our results indicate that ANNs can serve as a useful method to analyze quantitative effect of nsSNPs on protein function and would be useful for large-scale analysis of genomic nsSNP data. AVAILABILITY: The database is available for free at http://www.snp.mirworks.in.     

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.     

Partition dataset according to amino acid type improves the prediction of deleterious non-synonymous SNPs

Many non-synonymous SNPs (nsSNPs) are associated with diseases, and numerous machine learning methods have been applied to train classifiers for sorting disease-associated nsSNPs from neutral ones. The continuously accumulated nsSNP data allows us to further explore better prediction approaches. In this work, we partitioned the training data into 20 subsets according to either original or substituted amino acid type at the nsSNP site. Using support vector machine (SVM), training classification models on each subset resulted in an overall accuracy of 76.3% or 74.9% depending on the two different partition criteria, while training on the whole dataset obtained an accuracy of only 72.6%. Moreover, the dataset was also randomly divided into 20 subsets, but the corresponding accuracy was only 73.2%. Our results demonstrated that partitioning the whole training dataset into subsets properly, i.e., according to the residue type at the nsSNP site, will improve the performance of the trained classifiers significantly, which should be valuable in developing better tools for predicting the disease-association of nsSNPs.     

SNPdbe: constructing an nsSNP functional impacts database

Many existing databases annotate experimentally characterized single nucleotide polymorphisms (SNPs). Each non-synonymous SNP (nsSNP) changes one amino acid in the gene product (single amino acid substitution;SAAS). This change can either affect protein function or be neutral in that respect. Most polymorphisms lack experimental annotation of their functional impact. Here, we introduce SNPdbe-SNP database of effects, with predictions of computationally annotated functional impacts of SNPs. Database entries represent nsSNPs in dbSNP and 1000 Genomes collection, as well as variants from UniProt and PMD. SAASs come from >2600 organisms; 'human' being the most prevalent. The impact of each SAAS on protein function is predicted using the SNAP and SIFT algorithms and augmented with experimentally derived function/structure information and disease associations from PMD, OMIM and UniProt. SNPdbe is consistently updated and easily augmented with new sources of information. The database is available as an MySQL dump and via a web front end that allows searches with any combination of organism names, sequences and mutation IDs. AVAILABILITY: http://www.rostlab.org/services/snpdbe.     

In Silico profiling of deleterious amino acid substitutions of potential pathological importance in haemophlia A and haemophlia B

Background

In this study, instead of current biochemical methods, the effects of deleterious amino acid substitutions in F8 and F9 gene upon protein structure and function were assayed by means of computational methods and information from the databases. Deleterious substitutions of F8 and F9 are responsible for Haemophilia A and Haemophilia B which is the most common genetic disease of coagulation disorders in blood. Yet, distinguishing deleterious variants of F8 and F9 from the massive amount of nonfunctional variants that occur within a single genome is a significant challenge.

Methods

We performed an in silico analysis of deleterious mutations and their protein structure changes in order to analyze the correlation between mutation and disease. Deleterious nsSNPs were categorized based on empirical based and support vector machine based methods to predict the impact on protein functions. Furthermore, we modeled mutant proteins and compared them with the native protein for analysis of protein structure stability.

Results

Out of 510 nsSNPs in F8, 378 nsSNPs (74%) were predicted to be ''intolerant'' by SIFT, 371 nsSNPs (73%) were predicted to be ''damaging'' by PolyPhen and 445 nsSNPs (87%) as ''less stable'' by I-Mutant2.0. In F9, 129 nsSNPs (78%) were predicted to be intolerant by SIFT, 131 nsSNPs (79%) were predicted to be damaging by PolyPhen and 150 nsSNPs (90%) as less stable by I-Mutant2.0. Overall, we found that I-Mutant which emphasizes support vector machine based method outperformed SIFT and PolyPhen in prediction of deleterious nsSNPs in both F8 and F9.

Conclusions

The models built in this work would be appropriate for predicting the deleterious amino acid substitutions and their functions in gene regulation which would be useful for further genotype-phenotype researches as well as the pharmacogenetics studies. These in silico tools, despite being helpful in providing information about the nature of mutations, may also function as a first-pass filter to determine the substitutions worth pursuing for further experimental research in other coagulation disorder causing genes.     

SySAP: a system-level predictor of deleterious single amino acid polymorphisms

Single amino acid polymorphisms (SAPs), also known as non-synonymous single nucleotide polymorphisms (nsSNPs), are responsible for most of human genetic diseases. Discriminate the deleterious SAPs from neutral ones can help identify the disease genes and understand the mechanism of diseases. In this work, a method of deleterious SAP prediction at system level was established. Unlike most existing methods, our method not only considers the sequence and structure information, but also the network information. The integration of network information can improve the performance of deleterious SAP prediction. To make our method available to the public, we developed SySAP (a System-level predictor of deleterious Single Amino acid Polymorphisms), an easy-to-use and high accurate web server. SySAP is freely available at http://www.biosino.org/SySAP/and http://lifecenter.sgst.cn/SySAP/.     

Disulfide connectivity prediction using recursive neural networks and evolutionary information

MOTIVATION: We focus on the prediction of disulfide bridges in proteins starting from their amino acid sequence and from the knowledge of the disulfide bonding state of each cysteine. The location of disulfide bridges is a structural feature that conveys important information about the protein main chain conformation and can therefore help towards the solution of the folding problem. Existing approaches based on weighted graph matching algorithms do not take advantage of evolutionary information. Recursive neural networks (RNN), on the other hand, can handle in a natural way complex data structures such as graphs whose vertices are labeled by real vectors, allowing us to incorporate multiple alignment profiles in the graphical representation of disulfide connectivity patterns. RESULTS: The core of the method is the use of machine learning tools to rank alternative disulfide connectivity patterns. We develop an ad-hoc RNN architecture for scoring labeled undirected graphs that represent connectivity patterns. In order to compare our algorithm with previous methods, we report experimental results on the SWISS-PROT 39 dataset. We find that using multiple alignment profiles allows us to obtain significant prediction accuracy improvements, clearly demonstrating the important role played by evolutionary information. AVAILABILITY: The Web interface of the predictor is available at http://neural.dsi.unifi.it/cysteines     

DAMpred: Recognizing Disease-Associated nsSNPs through Bayes-Guided Neural-Network Model Built on Low-Resolution Structure Prediction of Proteins and Protein–Protein Interactions

Nearly one-third of non-synonymous single-nucleotide polymorphism (nsSNPs) are deleterious to human health, but recognition of the disease-associated mutations remains a significant unsolved problem. We proposed a new algorithm, DAMpred, to identify disease-causing nsSNPs through the coupling of evolutionary profiles with structure predictions of proteins and protein–protein interactions. The pipeline was trained by a novel Bayes-guided artificial neural network algorithm that incorporates posterior probabilities of distinct feature classifiers with the network training process. DAMpred was tested on a large-scale data set involving 10,635 nsSNPs from 2154 ORFs in the human genome and recognized disease-associated nsSNPs with an accuracy 0.80 and a Matthews correlation coefficient of 0.601, which is 9.1% higher than the best of other state-of-the-art methods. In the blind test on the TP53 gene, DAMpred correctly recognized the mutations causative of Li–Fraumeni-like syndrome with a Matthews correlation coefficient that is 27% higher than the control methods. The study demonstrates an efficient avenue to quantitatively model the association of nsSNPs with human diseases from low-resolution protein structure prediction, which should find important usefulness in diagnosis and treatment of genetic diseases.