Fuzzy neural network-based prediction of the motif for MHC class II binding peptides

Characterizing the interaction between major histocompatibility complex (MHC) molecules and antigenic peptides is critical for understanding immunity and developing immunotherapies for autoimmune diseases and cancer. To identify the peptide binding motif and predict peptides that bind to the human MHC classII molecule HLA-DR4(*0401), we applied a fuzzy neural network (FNN) capable of extracting the relationship between input and output. Analysis of the peptide binding motif revealed that the hydrophilicity of the position 1 residue located on the N-terminal side of the nonamer (9mer) was the most important variable and that the van der Waals volume and hydrophilicity of the position 6 residue and the hydrophilicity of the position 7 residue were also important variables. The estimation accuracy (A(ROC) value) was high and the binding motif extracted from the FNN agreed with that derived experimentally. This study demonstrates that FNN modeling allows candidate antigenic peptides to be selected without the need for further experiments.     

Construction of COD simulation model for activated sludge process by fuzzy neural network

Fuzzy neural network (FNN) was applied to construct a simulation model for estimating the effluent chemical oxygen demand (COD) value of an activated sludge process in a "U" plant, in which most of process variables were measured once an hour. The constructed FNN model could simulate periodic changes in COD with high accuracy. Comparing the simulation result obtained using the FNN model with that obtained using the multiple regression analysis (MRA) model, it was found that the FNN model had 3.7 times higher accuracy than the MRA model. The FNN models corresponding to each of the four seasons were also constructed. Analyzing the fuzzy rules acquired from the FNN models after learning, the operational characteristic of this plant could be elucidated. Construction of the simulation model for another plant "A", in which process variables were measured once a day, was also carried out. This FNN model also had a relatively high accuracy.     

Hidden Markov model-based prediction of antigenic peptides that interact with MHC class II molecules

Elucidating the interaction between major histocompatibility complex (MHC) molecules and antigenic peptides is fundamental to better understanding of the processes involved in immune responses and for the development of innovative immunotherapies. In the present study, hidden Markov models (HMM) were combined with the successive state splitting (SSS) algorithm for optimization of the HMM structure, to predict peptide binders to the human MHC class II molecule HLA-DRB1*0101. The predictive performance of our model (S-HMM) was compared with fully connected HMM and artificial neural network (ANN) methods using the relative operating characteristic (ROC) analysis. The S-HMM predictions had values of ROC > or = 0.85 which was at least as good, or better than the comparison methods. In addition, S-HMM is trained on positive data only and does not require exhaustive data preprocessing, such as peptide alignment. Our results demonstrated that S-HMM combines the high accuracy of predictions with the simplicity of implementation and is therefore useful for analyzing MHC class II binding peptides. In particular the S-HMM may be trained using only positive data and, the preprocessing of training data, such as peptide alignment and the selection of binding cores, is not required in this method.     

Prognostic predictor with multiple fuzzy neural models using expression profiles from DNA microarray for metastases of breast cancer

Gene expression profiling data from DNA microarray were analyzed using the fuzzy neural network (FNN) modeling method for predicting the distant metastases of breast cancer. The best model consisting of five genes was able to predict metastases of breast cancer with 94% accuracy. Furthermore, 100% accuracy was achieved by majoritarian decision using only 25 genes from five noninferior models which were constructed independently. From the constructed model, gene expression rules, which may cause distant metastases, were explicitly extracted and 60% of the metastases cases could be explained by this rule. The FNN modeling method described in this paper enables precise extraction of significant biological markers affecting prognosis without prior knowledge.     

Classification of intramural metastases and lymph node metastases of esophageal cancer from gene expression based on boosting and projective adaptive resonance theory

Esophageal cancer is a well-known cancer with poorer prognosis than other cancers. An optimal and individualized treatment protocol based on accurate diagnosis is urgently needed to improve the treatment of cancer patients. For this purpose, it is important to develop a sophisticated algorithm that can manage a large amount of data, such as gene expression data from DNA microarrays, for optimal and individualized diagnosis. Marker gene selection is essential in the analysis of gene expression data. We have already developed a combination method of the use of the projective adaptive resonance theory and that of a boosted fuzzy classifier with the SWEEP operator denoted PART-BFCS. This method is superior to other methods, and has four features, namely fast calculation, accurate prediction, reliable prediction, and rule extraction. In this study, we applied this method to analyze microarray data obtained from esophageal cancer patients. A combination method of PART-BFCS and the U-test was also investigated. It was necessary to use a specific type of BFCS, namely, BFCS-1,2, because the esophageal cancer data were very complexity. PART-BFCS and PART-BFCS with the U-test models showed higher performances than two conventional methods, namely, k-nearest neighbor (kNN) and weighted voting (WV). The genes including CDK6 could be found by our methods and excellent IF-THEN rules could be extracted. The genes selected in this study have a high potential as new diagnosis markers for esophageal cancer. These results indicate that the new methods can be used in marker gene selection for the diagnosis of cancer patients.     

A motif detection and classification method for peptide sequences using genetic programming

An exploration of common rules (property motifs) in amino acid sequences has been required for the design of novel sequences and elucidation of the interactions between molecules controlled by the structural or physical environment. In the present study, we developed a new method to search property motifs that are common in peptide sequence data. Our method comprises the following two characteristics: (i) the automatic determination of the position and length of common property motifs by calculating the physicochemical similarity of amino acids, and (ii) the quick and effective exploration of motif candidates that discriminates the positives and negatives by the introduction of genetic programming (GP). Our method was evaluated by two types of model data sets. First, the intentionally buried property motifs were searched in the artificially derived peptide data containing intentionally buried property motifs. As a result, the expected property motifs were correctly extracted by our algorithm. Second, the peptide data that interact with MHC class II molecules were analyzed as one of the models of biologically active peptides with buried motifs in various lengths. Twofold MHC class II binding peptides were identified with the rule using our method, compared to the existing scoring matrix method. In conclusion, our GP based motif searching approach enabled to obtain knowledge of functional aspects of the peptides without any prior knowledge.     

Selection of causal gene sets for lymphoma prognostication from expression profiling and construction of prognostic fuzzy neural network models

To assess the response of lymphomas to chemotherapy, gene expression profiling data from DNA microarrays were analyzed using the fuzzy neural network (FNN) modeling method. We used the FNN modeling method to produce 10 noninferior models. Using these models, we were able to predict diffuse large B-cell lymphoma (DLBCL) patient outcome with 93% accuracy. Of the 37 genes in the 10 models, 13 genes were repeatedly selected, indicating that these genes are important for prognostication. On Kaplan-Meier plots of overall survival, patients predicted by the FNN model to be cured survived significantly longer than those predicted to be refractory (P<0.0001), indicating that the FNN could successfully identify patients with a relatively poor prognosis among low-clinical-risk patients. The FNN modeling method presented here is able to precisely extract significant biological markers affecting prognosis.     

The class II major histocompatibility complex molecule BoLA-DR is expressed by endothelial cells of the bovine corpus luteum

Cells expressing class II major histocompatibility complex (MHC) molecules are found within the corpus luteum (CL) of several species. Expression and localization of class II MHC molecules in the bovine CL were examined in the present study. Immunohistochemical evaluation revealed class II MHC molecules on single cells in early CL (days 4 and 5 post-estrus). Two class II MHC-expressing cell types were observed in midcycle CL (days 10-12 post-estrus), single cells similar to those observed in the early CL, and endothelial cells. Not all endothelial cells expressed class II MHC, and further investigation revealed expression of only one type of class II MHC molecule, DR, on endothelial cells. Class II MHC was also localized to endothelial cells in late CL (day 18 post-estrus). Steroidogenic luteal cells were negative for class II MHC throughout the estrous cycle. Quantitative RT-PCR revealed higher (P < 0.05) concentrations of mRNA encoding the alpha-subunit of DR (DRA) in late CL when compared with those in the early CL. DRA mRNA abundance was also measured in cultures of mixed luteal and luteal endothelial (CLENDO) cells, in the presence or absence of tumor necrosis factor-alpha (TNF). No differences were found in the DRA mRNA concentration between mixed luteal and CLENDO cell cultures, and TNF had no effect on DRA mRNA concentration in both cell types. Expression of DR by endothelial cells of the midcycle CL may induce anergy of T lymphocytes, or stimulate them to secrete products that enhance normal luteal function.     

Negative energy balance does not decrease expression of leukocyte adhesion or antigen-presenting molecules in cattle

Sixteen yearling Holstein steers were fed for 210 or 60% of maintenance requirements to impose positive or negative energy balance, respectively. Blood was collected and analyzed for serum concentration of nonesterified fatty acids (NEFA), and leukocytes were isolated and counted. Isolated leukocytes were then analyzed for expression of the adhesion molecules L-selectin (CD62L), Mac-1 (CD11b and CD18), and major histocompatability complex (MHC) class I and class II molecules with immunostaining and flow cytometric analysis. Negative energy balance increased the concentration of NEFA in serum (P < 0.0001). Expression of CD62L on neutrophils was increased 14% during negative energy balance (P = 0.03). Energy balance did not affect expression of CD62L on any other cell types or expression of CD11b or CD18. Negative energy balance did not affect MHC class I expression but resulted in a small but significant increase in the expression of MHC class II (P = 0.03). The results of this study provide little evidence that nutritionally created negative energy balance impairs expression of CD62L, CD11b, and CD18 or expression of MHC class I or MHC class II molecules by resting bovine blood leukocytes.     

Sensory Modeling of Coffee with a Fuzzy Neural Network

ABSTRACT: Models were constructed to predict sensory evaluation scores from the blending ratio of coffee beans. Twenty-two blended coffees were prepared from 3 representative beans and were evaluated with respect to 10 sensory attributes by 5 coffee cup-tasters and by models constructed using the response surface method (RSM), multiple regression analysis (MRA), and a fuzzy neural network (FNN). The RSM and MRA models showed good correlations for some sensory attributes, but lacked sufficient overall accuracy. The FNN model exhibited high correlations for all attributes, clearly demonstrated the relationships between blending ratio and flavor characteristics, and was accurate enough for practical use. FNN, thus, constitutes a powerful tool for accelerating product development.     

Presence and regulation of messenger ribonucleic acids encoding components of the class II major histocompatibility complex-associated antigen processing pathway in the bovine corpus luteum

Luteal cells express class II major histocompatibility complex (MHC) molecules and can stimulate T lymphocyte proliferation in vitro. However, it is unknown whether luteal cells express the intracellular components necessary to process the peptides presented by class II MHC molecules. The objective of the present study was to examine the expression and regulation of three major class II-associated antigen processing components--class II MHC-associated invariant chain (Ii), DMalpha and DMbeta--in luteal tissue. Corpora lutea were collected early in the estrous cycle, during midcycle and late in the estrous cycle, and at various times following administration of a luteolytic dose of prostaglandin F(2alpha) (PGF(2alpha)) to the cow. Northern analysis revealed the presence of mRNA encoding each of the class II MHC-associated antigen processing proteins in luteal tissue. Ii mRNA concentrations did not change during the estrous cycle, whereas DMalpha and DMbeta mRNA concentrations were highest in midcycle luteal tissue compared with either early or late luteal tissue. Tumor necrosis factor-alpha (TNF-alpha) reduced DMalpha mRNA concentrations in cultured luteal cells in the presence of LH or PGF(2alpha). DMalpha and DMbeta mRNA were also present in highly enriched cultures of luteal endothelial (CLENDO) cells, and DMalpha mRNA concentrations were greater in CLENDO cultures compared with mixed luteal cell cultures. Expression of invariant chain, DMalpha and DMbeta genes indicates that cells within the corpus luteum express the minimal requirements to act as functional antigen-presenting cells, and the observation that CLENDO cells are a source of DMalpha and DMbeta mRNA indicates that non-immune cells within the corpus luteum may function as antigen-presenting cells.     

Screening of peptides with a high affinity to bile acids using peptide arrays and a computational analysis

Bile acid binding peptides have attracted attention for the improvement and prevention of hypercholesterolemia. In this study, screening of bile acid high affinity peptides was investigated using computationally-assisted peptide array analysis. Starting with the screening data obtained from a limited, random 6-mer library (2212 sequences), the peptides with a high affinity to bile acid were characterized by comparison of high- and low-affinity peptides using fuzzy neural network (FNN) analysis. The physical properties of amino acids at specific positions that contribute to bile acid binding activity were extracted as the structural rule; optimization was carried out using three repeated screening cycles of the rule extraction. The extracted structural rule indicates that Trp, Tyr, Phe, Leu, Ile and Val are enriched in bile acid binding peptides. The yields of bile acid binding peptides with an affinity of above the VAWWMY peptide (soystatin, control sequence) were significantly higher in the optimized structural rule (32.5%) compared to that of the random library (3.1%), and 6 peptides were obtained with above 2.0-fold increased binding activity.     

活性肽与MHC结合能力预测的免疫信息学方法研究进展

综述活性肽与MHC(主要组织相容性复合体,major histocompatibility complex)结合能力预测的免疫信息学方法的最新进展,介绍常用的肽与MHC分子结合预测的相关工具及方法,分析了各类方法的特点、研究重点和难点,以期为寻找免疫活性肽提供更快捷的方法。     

Short communication: Establishment of a new polymerase chain reaction-sequence-based typing method for genotyping cattle major histocompatibility complex class II DRB3

Sequence-based typing (SBT) is the most comprehensive method for characterizing major histocompatibility complex (MHC) gene polymorphisms. We report here a new PCR-SBT method for genotyping cattle MHC (BoLA) class II DRB3 using the Assign 400ATF ver. 1.0.2.41 software (Conexio Genomics, Fremantle, Australia), which detects alleles in a semiautomated manner. We examined 12 sets of PCR reactions for their ability to amplify BoLA-DRB3 exon 2 and selected an optimal primer set, which used ERB3N-HL031 for first-round PCR and ALL-DRB3B for second-round PCR. Next, we constructed a BoLA-DRB3 allele database using the reference sequences of the Assign 400ATF software and successfully assigned heterozygous samples (including those with deletion alleles) using bidirectional sequencing, unlike our previously described method, which used unidirectional sequencing for detecting of deletion alleles. Next, blood samples of 128 Holstein cattle were used to correlate the results of our modified PCR-SBT method with those of our previously described PCR-SBT method. Each new PCR-SBT result corresponded completely with the DRB3 allele that was genotyped by our previously described PCR-SBT method. Moreover, we confirmed the accuracy of our modified PCR-SBT method by genotyping 7 sire cattle and their 22 calves using Japanese Black cattle. This new method will contribute to high-throughput genotyping of BoLA-DRB3 by sequence-based typing.     

模糊权值网络最短路问题的数学模型及算法

模糊权值网络的最短路问题是一类重要的网络优化问题.针对边权值为三角模糊数的模糊权值网络的最短路问题,基于模糊数的结构元加权序,将其模糊线性规划模型等价转化为经典的线性规划模型,并提出一种改进的权矩阵算法来求解该问题,算法证明和应用实例表明新算法的正确性和有效性.此外对于边权值为其他形式模糊数的模糊权值网络的最短路问题,文中模型和算法同样有效.     

Forward modeling of metal complexation by NOM: II. prediction of binding site properties

An a priori model of metal complexation by natural organic matter (NOM) has previously been shown to predict experimental data at pH 7.0 and 0.1 M ionic strength (Cabaniss, S. E. Environ. Sci. Technol. 2009). Unlike macroscopic models based only on stoichiometry and thermodynamics, this a priori model also predicts the ligand groups and properties of complexed (occupied) molecules. Ligand molecules with strong binding sites form complexes at low metal concentrations and have average properties (molecular weight, charge, aromaticity) which can differ significantly from the average properties of bulk NOM. Cu(II), Ni(II) and Pb(II) preferentially bind to strong amine-containing sites which are often located on small (MW < 1000), lower-aromaticity molecules. Cd(II) and Zn(II) show generally weaker binding, although they also prefer amine-containing sites to pure carboxylates and bind to smaller, less aromatic molecules. Ca(II) shows no real preference for amine over carboxylate ligand groups, preferentially binding to larger and more negatively charged molecules. Al(III) has a unique preference for phenol-containing sites and larger, more aromatic molecules. While some predictions of this model are consistent with a variety of experimental data from the literature, others await validation by molecular-level analysis.     

NY‐ESO‐1 protein glycosylated by yeast induces enhanced immune responses

Vaccine strategies that target dendritic cells to elicit potent cellular immunity are the subject of intense research. Here we report that the genetically engineered yeast Saccharomyces cerevisiae, expressing the full‐length tumour‐associated antigen NY‐ESO‐1, is a versatile host for protein production. Exposing dendritic cells (DCs) to soluble NY‐ESO‐1 protein linked to the yeast a‐agglutinin 2 protein (Aga2p) protein resulted in protein uptake, processing and MHC class I cross‐presentation of NY‐ESO‐1‐derived peptides. The process of antigen uptake and cross‐presentation was dependent on the glycosylation pattern of NY‐ESO‐1‐Aga2p protein and the presence of accessible mannose receptors. In addition, NY‐ESO‐1‐Aga2p protein uptake by dendritic cells resulted in recognition by HLA‐DP4 NY‐ESO‐1‐specific CD4⁺ T cells, indicating MHC class II presentation. Finally, vaccination of mice with yeast‐derived NY‐ESO‐1‐Aga2p protein led to an enhanced humoral and cellular immune response, when compared to the bacterially expressed NY‐ESO‐1 protein. Together, these data demonstrate that yeast‐derived full‐length NY‐ESO‐1‐Aga2p protein is processed and presented efficiently by MHC class I and II complexes and warrants clinical trials to determine the potential value of S. cerevisiae as a host for cancer vaccine development. Copyright © 2010 John Wiley & Sons, Ltd.     

基于变论域模糊-Smith的白酒勾调控制

目的:满足白酒勾调控制系统响应无超调的要求,消除系统时滞性对其控制效果的影响。方法:提出了一种基于变论域模糊-Smith的白酒勾调控制算法。对系统的控制过程及特点进行分析并建立近似的数学模型;引用Smith预估器补偿系统存在的时滞性,利用模糊控制的特点解决系统数学模型建立不精确的问题,并且结合了变论域方法来消除常规模糊控制精度不高,存在稳态误差的弊端;通过Matlab实现了该算法的仿真。结果:与PID-Smith以及模糊-Smith控制算法相比,该算法在确保系统响应无超调的前提下,具备更快的调节速度以及鲁棒性。结论:变论域模糊-Smith算法能够实现白酒勾调控制系统对流量的精确控制。     

基于模糊控制技术的啤酒快速发酵控制

介绍了模糊控制技术在啤酒两罐法生产中的具体应用。此技术不仅解决了倒罐过程中的温度控制问题,同时也大大缩短了倒罐时间、降低了劳动强度、提高了产品质量。利用模糊控制技术建立受控系统的模糊模型,根据操作人员的操作经验,建立受控对象的模糊关系矩阵,进而实现模糊控制。为了寻求最佳控制效果,在应用过程中逐步对模糊关系矩阵进行修改,从而提高受控对象的精度。