Systematic tuning of parameters in support vector clustering

Clustering algorithms divide a set of observations into groups so that members of the same group share common features. In most of the algorithms, tunable parameters are set arbitrarily or by trial and error, resulting in less than optimal clustering. This paper presents a global optimization strategy for the systematic and optimal selection of parameter values associated with a clustering method. In the process, a performance criterion for the optimization model is proposed and benchmarked against popular performance criteria from the literature (namely, the Silhouette coefficient, Dunn's index, and Davies-Bouldin index). The tuning strategy is illustrated using the support vector clustering (SVC) algorithm and simulated annealing. In order to reduce the computational burden, the paper also proposes an alternative to the adjacency matrix method (used for the assignment of cluster labels), namely the contour plotting approach. Datasets tested include the iris and the thyroid datasets from the UCI repository, as well as lymphoma and breast cancer data. The optimal tuning parameters are determined efficiently, while the contour plotting approach leads to significant reductions in computational effort (CPU time) especially for large datasets. The performance criteria comparisons indicate mixed results. Specifically, the Silhouette coefficient and the Davies-Bouldin index perform better, while the Dunn's index is worse on average than the proposed performance index.     

sRNASVM——基于SVM方法构建大肠杆菌sRNA预测模型

在理解细菌与环境的相互作用方面,细菌sRNA的识别发挥重要作用。文章介绍了一个通过增加训练集中实验证实的sRNA来构建细菌sRNA预测模型的策略,并以大肠杆菌K-12的sRNA预测为例来说明策略的可行性。结果表明,按此策略构建的模型sRNASVM的10倍交叉检验精度达到92.45%,高于目前文献中报道的精度。因此,构建的这一模型将为实验发现sRNA提供较好的生物信息学支持。有关模型和详细结果可以从网站http://ccb.bmi.ac.cn/srnasvm/下载。     

Prediction of protein structural class for low-similarity sequences using support vector machine and PSI-BLAST profile

Knowledge of structural class plays an important role in understanding protein folding patterns. In this study, a simple and powerful computational method, which combines support vector machine with PSI-BLAST profile, is proposed to predict protein structural class for low-similarity sequences. The evolution information encoding in the PSI-BLAST profiles is converted into a series of fixed-length feature vectors by extracting amino acid composition and dipeptide composition from the profiles. The resulting vectors are then fed to a support vector machine classifier for the prediction of protein structural class. To evaluate the performance of the proposed method, jackknife cross-validation tests are performed on two widely used benchmark datasets, 1189 (containing 1092 proteins) and 25PDB (containing 1673 proteins) with sequence similarity lower than 40% and 25%, respectively. The overall accuracies attain 70.7% and 72.9% for 1189 and 25PDB datasets, respectively. Comparison of our results with other methods shows that our method is very promising to predict protein structural class particularly for low-similarity datasets and may at least play an important complementary role to existing methods.     

Prediction of prostate cancer using hair trace element concentration and support vector machine method

A change in the normal concentration of essential trace elements in the human body might lead to major health disturbances. In this study, hair samples were collected from 115 human subject, including 55 healthy people and 60 patients with prostate cancer. The concentrations of 20 trace elements (TEs) in these samples were measured by inductively coupled plasma-mass spectrometry. A support vector machine was used to investigate the relationship between TEs and prostate cancer. It is found that, among the 20 TEs, 10 (Mg P, K, Ca, Cr, Mn, Fe. Cu, Zn, and Se) are related to the risk of prostate cancer. These 10 TEs were used to build the prediction model for prostate cancer. The model obtained can satisfactorily distinguish the healthy samples from the cancer samples. Furthermore, the cross-validation by leaving-one method proved that the prediction ability of this model reaches as high as 95.8%. It is practical to predict the risk of prostate cancer using this model in the clinics     

A novel representation for apoptosis protein subcellular localization prediction using support vector machine

Apoptosis, or programmed cell death, plays an important role in development of an organism. Obtaining information on subcellular location of apoptosis proteins is very helpful to understand the apoptosis mechanism. In this paper, based on the concept that the position distribution information of amino acids is closely related with the structure and function of proteins, we introduce the concept of distance frequency [Matsuda, S., Vert, J.P., Ueda, N., Toh, H., Akutsu, T., 2005. A novel representation of protein sequences for prediction of subcellular location using support vector machines. Protein Sci. 14, 2804-2813] and propose a novel way to calculate distance frequencies. In order to calculate the local features, each protein sequence is separated into p parts with the same length in our paper. Then we use the novel representation of protein sequences and adopt support vector machine to predict subcellular location. The overall prediction accuracy is significantly improved by jackknife test.     

复杂疾病驱使的融合SDA-SVM集成基因挖掘方法

提出了一种新颖的复杂疾病驱使的融合SDA-SVM(Stepwise Discriminant Analysis-Support Vector Machine,SDA-SVM)技术的集成基因挖掘方法。该集成方法融合逐步判别分析和支持向量机的优点,能够有效地进行复杂疾病相关基因的深度挖掘,使得挖掘出的基因能够较好地识别疾病类型和亚型。通过将该方法应用于一套弥散性大B细胞淋巴瘤DNA表达谱数据,并与其它基因挖掘方法对比,结果表明该方法挖掘出的基因具有较高的疾病相关性和较强的疾病类型识别能力。     

Prediction of protein structural classes using support vector machines

Summary. The support vector machine, a machine-learning method, is used to predict the four structural classes, i.e. mainly α, mainly β, α–β and fss, from the topology-level of CATH protein structure database. For the binary classification, any two structural classes which do not share any secondary structure such as α and β elements could be classified with as high as 90% accuracy. The accuracy, however, will decrease to less than 70% if the structural classes to be classified contain structure elements in common. Our study also shows that the dimensions of feature space 20² = 400 (for dipeptide) and 20³ = 8 000 (for tripeptide) give nearly the same prediction accuracy. Among these 4 structural classes, multi-class classification gives an overall accuracy of about 52%, indicating that the multi-class classification technique in support of vector machines may still need to be further improved in future investigation.     

Accuracy of dose calculation on iterative CBCT for head and neck radiotherapy

PurposeTo evaluate the feasibility of the use of iterative cone-beam computed tomography (CBCT) for dose calculation in the head and neck region.MethodsThis study includes phantom and clinical studies. All acquired CBCT images were reconstructed with Feldkamp–Davis–Kress algorithm-based CBCT (FDK-CBCT) and iterative CBCT (iCBCT) algorithm. The Hounsfield unit (HU) consistency between the head and body phantoms was determined in both reconstruction techniques. Volumetric modulated arc therapy (VMAT) plans were generated for 16 head and neck patients on a planning CT scan, and the doses were recalculated on FDK-CBCT and iCBCT with Anisotropic Analytical Algorithm (AAA) and Acuros XB (AXB). As a comparison of the accuracy of dose calculations, the absolute dosimetric difference and 1%/1 mm gamma passing rate analysis were analyzed.ResultsThe difference in the mean HU values between the head and body phantoms was larger for FDK-CBCT (max value: 449.1 HU) than iCBCT (260.0 HU). The median dosimetric difference from the planning CT were <1.0% for both FDK-CBCT and iCBCT but smaller differences were found with iCBCT (planning target volume D_50%: 0.38% (0.15–0.59%) for FDK-CBCT, 0.28% (0.13–0.49%) for iCBCT, AAA; 0.14% (0.04–0.19%) for FDK-CBCT, 0.07% (0.02–0.20%) for iCBCT). The mean gamma passing rate was significantly better in iCBCT than FDK-CBCT (AAA: 98.7% for FDK-CBCT, 99.4% for iCBCT; AXB: 96.8% for FDK_CBCT, 97.5% for iCBCT).ConclusionThe iCBCT-based dose calculation in VMAT for head and neck cancer was accurate compared to FDK-CBCT.     

Using Chou's amphiphilic pseudo-amino acid composition and support vector machine for prediction of enzyme subfamily classes

With the rapid increment of protein sequence data, it is indispensable to develop automated and reliable predictive methods for protein function annotation. One approach for facilitating protein function prediction is to classify proteins into functional families from primary sequence. Being the most important group of all proteins, the accurate prediction for enzyme family classes and subfamily classes is closely related to their biological functions. In this paper, for the prediction of enzyme subfamily classes, the Chou's amphiphilic pseudo-amino acid composition [Chou, K.C., 2005. Using amphiphilic pseudo amino acid composition to predict enzyme subfamily classes. Bioinformatics 21, 10-19] has been adopted to represent the protein samples for training the 'one-versus-rest' support vector machine. As a demonstration, the jackknife test was performed on the dataset that contains 2640 oxidoreductase sequences classified into 16 subfamily classes [Chou, K.C., Elrod, D.W., 2003. Prediction of enzyme family classes. J. Proteome Res. 2, 183-190]. The overall accuracy thus obtained was 80.87%. The significant enhancement in the accuracy indicates that the current method might play a complementary role to the exiting methods.     

Prediction of biological activity of Aurora-A kinase inhibitors by multilinear regression analysis and support vector machine

Several QSAR (quantitative structure-activity relationships) models for predicting the inhibitory activity of 117 Aurora-A kinase inhibitors were developed. The whole dataset was split into a training set and a test set based on two different methods, (1) by a random selection; and (2) on the basis of a Kohonen’s self-organizing map (SOM). Then the inhibitory activity of 117 Aurora-A kinase inhibitors was predicted using multilinear regression (MLR) analysis and support vector machine (SVM) methods, respectively. For the two MLR models and the two SVM models, for the test sets, the correlation coefficients of over 0.92 were achieved.     

Pseudo amino acid composition and multi-class support vector machines approach for conotoxin superfamily classification

Conotoxins are disulfide rich small peptides that target a broad spectrum of ion-channels and neuronal receptors. They offer promising avenues in the treatment of chronic pain, epilepsy and cardiovascular diseases. Assignment of newly sequenced mature conotoxins into appropriate superfamilies using a computational approach could provide valuable preliminary information on the biological and pharmacological functions of the toxins. However, creation of protein sequence patterns for the reliable identification and classification of new conotoxin sequences may not be effective due to the hypervariability of mature toxins. With the aim of formulating an in silico approach for the classification of conotoxins into superfamilies, we have incorporated the concept of pseudo-amino acid composition to represent a peptide in a mathematical framework that includes the sequence-order effect along with conventional amino acid composition. The polarity index attribute, which encodes information such as residue surface buriability, polarity, and hydropathy, was used to store the sequence-order effect. Several methods like BLAST, ISort (Intimate Sorting) predictor, least Hamming distance algorithm, least Euclidean distance algorithm and multi-class support vector machines (SVMs), were explored for superfamily identification. The SVMs outperform other methods providing an overall accuracy of 88.1% for all correct predictions with generalized squared correlation of 0.75 using jackknife cross-validation test for A, M, O and T superfamilies and a negative set consisting of short cysteine rich sequences from different eukaryotes having diverse functions. The computed sensitivity and specificity for the superfamilies were found to be in the range of 84.0-94.1% and 80.0-95.5%, respectively, attesting to the efficacy of multi-class SVMs for the successful in silico classification of the conotoxins into their superfamilies.     

Prediction of siRNA functionality using generalized string kernel and support vector machine

Small interfering RNAs (siRNAs) are becoming widely used for sequence-specific gene silencing in mammalian cells, but designing an effective siRNA is still a challenging task. In this study, we developed an algorithm for predicting siRNA functionality by using generalized string kernel (GSK) combined with support vector machine (SVM). With GSK, siRNA sequences were represented as vectors in a multi-dimensional feature space according to the numbers of subsequences in each siRNA, and subsequently classified with SVM into effective or ineffective siRNAs. We applied this algorithm to published siRNAs, and could classify effective and ineffective siRNAs with 90.6%, 86.2% accuracy, respectively.     

AutoMotif Server for prediction of phosphorylation sites in proteins using support vector machine: 2007 update

We present here the recent update of AutoMotif Server (AMS 2.0) that predicts post-translational modification sites in protein sequences. The support vector machine (SVM) algorithm was trained on data gathered in 2007 from various sets of proteins containing experimentally verified chemical modifications of proteins. Short sequence segments around a modification site were dissected from a parent protein, and represented in the training set as binary or profile vectors. The updated efficiency of the SVM classification for each type of modification and the predictive power of both representations were estimated using leave-one-out tests for model of general phosphorylation and for modifications catalyzed by several specific protein kinases. The accuracy of the method was improved in comparison to the previous version of the service (Plewczynski et al., “AutoMotif server: prediction of single residue post-translational modifications in proteins”, Bioinformatics 21: 2525–7, 2005). The precision of the updated version reached over 90% for selected types of phosphorylation and was optimized in trade of lower recall value of the classification model. The AutoMotif Server version 2007 is freely available at . Additionally, the reference dataset for optimization of prediction of phosphorylation sites, collected from the UniProtKB was also provided and can be accessed at .     

Identification of the functional alteration signatures across different cancer types with support vector machine and feature analysis

Cancers are regarded as malignant proliferations of tumor cells present in many tissues and organs, which can severely curtail the quality of human life. The potential of using plasma DNA for cancer detection has been widely recognized, leading to the need of mapping the tissue-of-origin through the identification of somatic mutations. With cutting-edge technologies, such as next-generation sequencing, numerous somatic mutations have been identified, and the mutation signatures have been uncovered across different cancer types. However, somatic mutations are not independent events in carcinogenesis but exert functional effects. In this study, we applied a pan-cancer analysis to five types of cancers: (I) breast cancer (BRCA), (II) colorectal adenocarcinoma (COADREAD), (III) head and neck squamous cell carcinoma (HNSC), (IV) kidney renal clear cell carcinoma (KIRC), and (V) ovarian cancer (OV). Based on the mutated genes of patients suffering from one of the aforementioned cancer types, patients they were encoded into a large number of numerical values based upon the enrichment theory of gene ontology (GO) terms and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. We analyzed these features with the Monte-Carlo Feature Selection (MCFS) method, followed by the incremental feature selection (IFS) method to identify functional alteration features that could be used to build the support vector machine (SVM)-based classifier for distinguishing the five types of cancers. Our results showed that the optimal classifier with the selected 344 features had the highest Matthews correlation coefficient value of 0.523. Sixteen decision rules produced by the MCFS method can yield an overall accuracy of 0.498 for the classification of the five cancer types. Further analysis indicated that some of these features and rules were supported by previous experiments. This study not only presents a new approach to mapping the tissue-of-origin for cancer detection but also unveils the specific functional alterations of each cancer type, providing insight into cancer-specific functional aberrations as potential therapeutic targets. This article is part of a Special Issue entitled: Accelerating Precision Medicine through Genetic and Genomic Big Data Analysis edited by Yudong Cai & Tao Huang.     

基于多分类支持向量机的基因表达系列分析

基因表达系列分析(Serial analysis of gene expression,SAGE)是一种基因表达数据,反映了细胞内的动态变化。模式识别和可视化方法是分析SAGE数据的基本工具,但是由于缺乏描述数据的统计特性,传统的聚类分析技术不适用于SAGE数据的分析。本文提出了一种基于多分类和支持向量机的SAGE数据的分析法。经过对模拟数据和人类癌症SAGE数据的分析,基于径向基核函数的多分类支持向量机算法一对一(one-against-one,OAO)算法提供了比PoissonC和PoissonS更好的分类结果。