Using stacked generalization to predict membrane protein types based on pseudo-amino acid composition

Membrane proteins are vitally important for many biological processes and have become an attractive target for both basic research and drug design. Knowledge of membrane protein types often provides useful clues in deducing the functions of uncharacterized membrane proteins. With the unprecedented increasing of newly found protein sequences in the post-genomic era, it is highly demanded to develop an automated method for fast and accurately identifying the types of membrane proteins according to their amino acid sequences. Although quite a few identifiers have been developed in this regard through various approaches, such as covariant discriminant (CD), support vector machine (SVM), artificial neural network (ANN), and K-nearest neighbor (KNN), classifier the way they operate the identification is basically individual. As is well known, wise persons usually take into account the opinions from several experts rather than rely on only one when they are making critical decisions. Likewise, a sophisticated identifier should be trained by several different modes. In view of this, based on the frame of pseudo-amino acid that can incorporate a considerable amount of sequence-order effects, a novel approach called "stacked generalization" or "stacking" has been introduced. Unlike the "bagging" and "boosting" approaches which only combine the classifiers of a same type, the stacking approach can combine several different types of classifiers through a meta-classifier to maximize the generalization accuracy. The results thus obtained were very encouraging. It is anticipated that the stacking approach may also hold a high potential to improve the identification quality for, among many other protein attributes, subcellular location, enzyme family class, protease type, and protein-protein interaction type. The stacked generalization classifier is available as a web-server named "SG-MPt_Pred" at: http://202.120.37.186/bioinf/wangsq/service.htm.     

A non-linear model for measuring grapevine leaf thickness by means of red-edge/near-infrared spectral reflectance

Vegetation is a key element of our ecology system. The leaf area and its thickness provide valuable information about the status of our environment. Thus, there is a need for accurate, efficient, practical methodologies to estimate this biochemical parameter. Hyperspectral measurement is a means of quickly assessing leaf parameter in situ. In the past decades, there were lots of work (Boyd et al.) that focused on measurement of leaf area index, but very few on measurement of leaf thickness. In this paper, reflectance of grape leaves was measured over the spectral range of 350–1010 nm. The corresponding thickness of leaves from four grapevine cultivars was also measured as part of seventeen field campaigns undertaken during the summer of 2007. An artificial-intelligence technique, the support vector machine (SVM) model, was introduced to establish the relationship between the leaf thickness and red-edge/near-infrared (NIR) reflectance, with variability examined among individual cultivars as well as at various growth stages. The best wavelengths were variable depending on the grape cultivar and growth stage. The SVM model allows compilation of factors such as cultivar and growth stage with spectral information to yield a superior result.     

Differential pulse voltammetric simultaneous determination of acetaminophen and ascorbic acid using single-walled carbon nanotube-modified carbon-ceramic electrode

Single-walled carbon nanotube-modified carbon–ceramic electrode (SWCNT/CCE) was employed for the simultaneous determination of acetaminophen (APAP) and ascorbic acid (AA). The SWCNT/CCE displayed excellent electrochemical catalytic activities toward APAP and AA oxidation compared with bare CCE. In the differential pulse voltammetry technique, both AA and APAP gave sensitive oxidation peaks at −62 and 302 mV versus saturated calomel electrode, respectively. Under the optimized experimental conditions, APAP and AA gave linear responses over ranges of 0.2 to 150.0 μM (R² = 0.998) and 5.0 to 700.0 μM (R² = 0.992), respectively. The lower detection limits were found to be 0.12 μM for APAP and 3.0 μM for AA. The investigated method showed good stability, reproducibility, and repeatability as well as high recovery in pharmaceutical and biological samples.     

Prediction-based structured variable selection through the receiver operating characteristic curves

In many clinical settings, a commonly encountered problem is to assess accuracy of a screening test for early detection of a disease. In these applications, predictive performance of the test is of interest. Variable selection may be useful in designing a medical test. An example is a research study conducted to design a new screening test by selecting variables from an existing screener with a hierarchical structure among variables: there are several root questions followed by their stem questions. The stem questions will only be asked after a subject has answered the root question. It is therefore unreasonable to select a model that only contains stem variables but not its root variable. In this work, we propose methods to perform variable selection with structured variables when predictive accuracy of a diagnostic test is the main concern of the analysis. We take a linear combination of individual variables to form a combined test. We then maximize a direct summary measure of the predictive performance of the test, the area under a receiver operating characteristic curve (AUC of an ROC), subject to a penalty function to control for overfitting. Since maximizing empirical AUC of the ROC of a combined test is a complicated nonconvex problem (Pepe, Cai, and Longton, 2006, Biometrics62, 221-229), we explore the connection between the empirical AUC and a support vector machine (SVM). We cast the problem of maximizing predictive performance of a combined test as a penalized SVM problem and apply a reparametrization to impose the hierarchical structure among variables. We also describe a penalized logistic regression variable selection procedure for structured variables and compare it with the ROC-based approaches. We use simulation studies based on real data to examine performance of the proposed methods. Finally we apply developed methods to design a structured screener to be used in primary care clinics to refer potentially psychotic patients for further specialty diagnostics and treatment.     

Sox2 and Fgf interact with Atoh1 to promote sensory competence throughout the zebrafish inner ear

Atoh1 is required for differentiation of sensory hair cells in the vertebrate inner ear. Moreover, misexpression of Atoh1 is sufficient to establish ectopic sensory epithelia, making Atoh1 a good candidate for gene therapy to restore hearing. However, competence to form sensory epithelia appears to be limited to discrete regions of the inner ear. To better understand the developmental factors influencing sensory-competence, we examined the effects of misexpressing atoh1a in zebrafish embryos under various developmental conditions. Activation of a heat shock-inducible transgene, hs:atoh1a, resulted in ectopic expression of early markers of sensory development within 2 h, and mature hair cells marked by brn3c:GFP began to accumulate 9 h after heat shock. The ability of atoh1a to induce ectopic sensory epithelia was maximal when activated during placodal or early otic vesicle stages but declined rapidly thereafter. At no stage was atoh1a sufficient to induce sensory development in dorsal or lateral non-sensory regions of the otic vesicle. However, co-misexpression of atoh1a with fgf3, fgf8 or sox2, genes normally acting in the same gene network with atoh1a, stimulated sensory development in all regions of the otic vesicle. Thus, expression of fgf3, fgf8 or sox2 strongly enhances competence to respond to Atoh1.     

基于Z曲线方法重新注释脑膜炎奈瑟菌MC58株基因组

已测序的微生物基因组中包含的注释开放阅读框(open reading frames,ORFs)可以分为两大类:第一类对应于功能已知的蛋白质编码基因;第二类则为功能未知的假设ORFs,其中通常有一部分实际上不编码蛋白质。采用基于Z曲线的方法从属于第一类的功能已知基因出发训练参数,进而确定第二类ORFs中非编码的部分。通过支持向量机的学习及分类,结果显示十重交叉检验平均正确率为98.45%,说明Z曲线联合支持向量机是一种高度准确的基因识别方法。最终,确定216个假设ORFs实际上不编码蛋白质。通过采用Blastp进行序列比对,保留的假设ORFs中有341个在高可靠性的条件下获得了功能信息。根据蛋白质直系同源簇方法进行功能分类,分别有30、53、59和159个新注释的假设ORFs属于信息储存和加工类、细胞加工和信号传递类、新陈代谢类和特征不明显类。另外还有70个不属于其中的任何一类。注释结果比RefSeq及GenBank提供的原注释更加准确,更加完整。     

基于支持向量机和贝叶斯方法的蛋白质四级结构分类研究

用支持向量机和贝叶斯两种方法对蛋白质四级结构进行分类研究。结果表明,基于支持向量机的分类结果最好,其l0CV检验的总分类精度、正样本正确预测率、Matthes相关系数和假阳性率分别为74．2％、84．6％、0．474、38．9％;基于贝叶斯的分类结果没有支持向量机的分类结果好,但其l0CV检验的假阳性率最低(15．9％)．这些结果说明同源寡聚蛋白质一级序列包含四级结构信息,同时特征向量的确表示了埋藏在缔合亚基作用部位接触表面的基本信息。     

基于支持向量机方法的蛋白可溶性预测

按照蛋白质序列中残基的相对可溶性,将其分为两类(表面／内部)和三类(表面／中间／内部)进行预测。选择不同窗宽和参数对数据进行训练和预测,以确保得到最好的分类效果,并同其他已有方法进行比较。对同一数据集不同分类阈值的预测结果显示,支持向量机方法对蛋白质可溶性的整体预测效果优于神经网络和信息论的方法。其中,对两类数据的最优分类结果达到79．0％,对三类数据的最优分类结果达到67．5％,表明支持向量机是蛋白质残基可溶性预测的一种有效方法。     

对应分析在近红外有效波长选择的应用

针对分立波长型近红外判别仪器开发过程中遇到的组合波长优选问题,提出一种基于对应分析的有效波长选择方法.对11种蘑菇样品及其在701-2500nm的光谱信息作对应分析,在对应图上计算各样品点与波长点的距离后,首先由各样品点的最近距离波长点确定出最稳健波长组合,在此基础上,对各样品点的前10、50、200个最近距离波长点进行优选,得出用921nm、1376nm、1424nm、1720nm、2233nm、2454nm这6个波长建立的SVM模型,能在输入变量最少的情况下达到用最稳健波长组合所建模型的预测能力.     

Using pseudo-amino acid composition and support vector machine to predict protein structural class

As a result of genome and other sequencing projects, the gap between the number of known protein sequences and the number of known protein structural classes is widening rapidly. In order to narrow this gap, it is vitally important to develop a computational prediction method for fast and accurately determining the protein structural class. In this paper, a novel predictor is developed for predicting protein structural class. It is featured by employing a support vector machine learning system and using a different pseudo-amino acid composition (PseAA), which was introduced to, to some extent, take into account the sequence-order effects to represent protein samples. As a demonstration, the jackknife cross-validation test was performed on a working dataset that contains 204 non-homologous proteins. The predicted results are very encouraging, indicating that the current predictor featured with the PseAA may play an important complementary role to the elegant covariant discriminant predictor and other existing algorithms.