Genome-enabled prediction using probabilistic neural network classifiers.

BACKGROUND: Multi-layer perceptron (MLP) and radial basis function neural networks (RBFNN) have been shown to be effective in genome-enabled prediction. Here, we evaluated and compared the classification performance of an MLP classifier versus that of a probabilistic neural network (PNN), to predict the probability of membership of one individual in a phenotypic class of interest, using genomic and phenotypic data as input variables. We used 16 maize and 17 wheat genomic and phenotypic datasets with different trait-environment combinations (sample sizes ranged from 290 to 300 individuals) with 1.4 k and 55 k SNP chips. Classifiers were tested using continuous traits that were categorized into three classes (upper, middle and lower) based on the empirical distribution of each trait, constructed on the basis of two percentiles (15-85 % and 30-70 %). We focused on the 15 and 30 % percentiles for the upper and lower classes for selecting the best individuals, as commonly done in genomic selection. Wheat datasets were also used with two classes. The criteria for assessing the predictive accuracy of the two classifiers were the area under the receiver operating characteristic curve (AUC) and the area under the precision-recall curve (AUCpr). Parameters of both classifiers were estimated by optimizing the AUC for a specific class of interest. RESULTS: The AUC and AUCpr criteria provided enough evidence to conclude that PNN was more accurate than MLP for assigning maize and wheat lines to the correct upper, middle or lower class for the complex traits analyzed. Results for the wheat datasets with continuous traits split into two and three classes showed that the performance of PNN with three classes was higher than with two classes when classifying individuals into the upper and lower (15 or 30 %) categories. CONCLUSIONS: The PNN classifier outperformed the MLP classifier in all 33 (maize and wheat) datasets when using AUC and AUCpr for selecting individuals of a specific class. Use of PNN with Gaussian radial basis functions seems promising in genomic selection for identifying the best individuals. Categorizing continuous traits into three classes generally provided better classification than when using two classes, because classification accuracy improved when classes were balanced.

Characterization of a variant of Xanthomonas citri subsp. citri that triggers a host-specific defense response.

Citrus is an economically important fruit crop that is severely afflicted by Asiatic citrus bacterial canker (CBC), a disease caused by the phytopathogen Xanthomonas citri subsp. citri (X. citri). To gain insight into the molecular epidemiology of CBC, 42 Xanthomonas isolates were collected from a range of Citrus spp. across 17 different orchards in Tucumán, Argentina and subjected to molecular, biochemical, and pathogenicity tests. Analysis of genome-specific X. citri markers and DNA polymorphisms based on repetitive elements-based polymerase chain reaction showed that all 42 isolates belonged to X. citri. Interestingly, pathogenicity tests showed that one isolate, which shares >90% genetic similarity to the reference strain X. citri T, has host range specificity. This new variant of X. citri subsp. citri, named X. citri A(T), which is deficient in xanthan production, induces an atypical, noncankerous chlorotic phenotype in Citrus limon and C. paradisi and weak cankerous lesions in C. aurantifolia and C. clementina leaves. In C. limon, suppression of canker development is concomitant with an oxidative burst; xanthan is not implicated in the phenotype induced by this interaction, suggesting that other bacterial factors would be involved in triggering the defense response.

Multiclass classification of microarray data samples with a reduced number of genes.

BACKGROUND: Multiclass classification of microarray data samples with a reduced number of genes is a rich and challenging problem in Bioinformatics research. The problem gets harder as the number of classes is increased. In addition, the performance of most classifiers is tightly linked to the effectiveness of mandatory gene selection methods. Critical to gene selection is the availability of estimates about the maximum number of genes that can be handled by any classification algorithm. Lack of such estimates may lead to either computationally demanding explorations of a search space with thousands of dimensions or classification models based on gene sets of unrestricted size. In the former case, unbiased but possibly overfitted classification models may arise. In the latter case, biased classification models unable to support statistically significant findings may be obtained. RESULTS: A novel bound on the maximum number of genes that can be handled by binary classifiers in binary mediated multiclass classification algorithms of microarray data samples is presented. The bound suggests that high-dimensional binary output domains might favor the existence of accurate and sparse binary mediated multiclass classifiers for microarray data samples. CONCLUSIONS: A comprehensive experimental work shows that the bound is indeed useful to induce accurate and sparse multiclass classifiers for microarray data samples.

Applications of Machine Learning Methods to Genomic Selection in Breeding Wheat for Rust Resistance.

New methods and algorithms are being developed for predicting untested phenotypes in schemes commonly used in genomic selection (GS). The prediction of disease resistance in GS has its own peculiarities: a) there is consensus about the additive nature of quantitative adult plant resistance (APR) genes, although epistasis has been found in some populations; b) rust resistance requires effective combinations of major and minor genes; and c) disease resistance is commonly measured based on ordinal scales (e.g., scales from 1-5, 1-9, etc.). Machine learning (ML) is a field of computer science that uses algorithms and existing samples to capture characteristics of target patterns. In this paper we discuss several state-of-the-art ML methods that could be applied in GS. Many of them have already been used to predict rust resistance in wheat. Others are very appealing, given their performance for predicting other wheat traits with similar characteristics. We briefly describe the proposed methods in the Appendix.

Applications of Genomic Selection in Breeding Wheat for Rust Resistance.

There are a lot of methods developed to predict untested phenotypes in schemes commonly used in genomic selection (GS) breeding. The use of GS for predicting disease resistance has its own particularities: (a) most populations shows additivity in quantitative adult plant resistance (APR); (b) resistance needs effective combinations of major and minor genes; and (c) phenotype is commonly expressed in ordinal categorical traits, whereas most parametric applications assume that the response variable is continuous and normally distributed. Machine learning methods (MLM) can take advantage of examples (data) that capture characteristics of interest from an unknown underlying probability distribution (i.e., data-driven). We introduce some state-of-the-art MLM capable to predict rust resistance in wheat. We also present two parametric R packages for the reader to be able to compare.

Development and evaluation of a colorimetric PCR system for the detection and typing of human papillomaviruses.

In developing countries, the introduction of human papillomaviruses (HPV) DNA testing as an adjunct to cytological screening programs has been delayed due to the lack of high performance and cost effective diagnostic nucleic acid methods. In this study we report the development and evaluation of the L1HPVPCR, a PCR-based method for the detection and typing of five of the most prevalent high-risk HPV types. The L1HPVPCR assay combines amplification with the MY09/11 HPV consensus primer system, liquid hybridization of the PCR products with no radioactive probes and enzyme immunoassay analysis. The technique is a user-friendly system that allows accurate HPV DNA detection and typing with inexpensive instrumentation that could be performed with not sophisticated reagents in almost any laboratory. Different cutoff points for generic and specific HPV detection were determined using reproducibility analysis and receiver operating characteristic curves to ensure good analytical sensitivity and clinical effectiveness. We used the L1HPVPCR assay to estimate the prevalence of HPV infection in 127 women at risk of cervical cancer from the city of Rosario (Argentina), where no epidemiological data has been previously reported. Further, we explored the clinical utility of the L1HPVPCR assay respect the Pap smear using a combined diagnosis of cytology, histology and colposcopy as gold standard. In conclusion, our results indicate that the assay described here provides a tool for accurate HPV DNA testing and could be applied in regions where no commercial tests are available.

Reference genes for real-time PCR quantification of microRNAs and messenger RNAs in rat models of hepatotoxicity.

Hepatotoxicity is associated with major changes in liver gene expression induced by xenobiotic exposure. Understanding the underlying mechanisms is critical for its clinical diagnosis and treatment. MicroRNAs are key regulators of gene expression that control mRNA stability and translation, during normal development and pathology. The canonical technique to measure gene transcript levels is Real-Time qPCR, which has been successfully modified to determine the levels of microRNAs as well. However, in order to obtain accurate data in a multi-step method like RT-qPCR, the normalization with endogenous, stably expressed reference genes is mandatory. Since the expression stability of candidate reference genes varies greatly depending on experimental factors, the aim of our study was to identify a combination of genes for optimal normalization of microRNA and mRNA qPCR expression data in experimental models of acute hepatotoxicity. Rats were treated with four traditional hepatotoxins: acetaminophen, carbon tetrachloride, D-galactosamine and thioacetamide, and the liver expression levels of two groups of candidate reference genes, one for microRNA and the other for mRNA normalization, were determined by RT-qPCR in compliance with the MIQE guidelines. In the present study, we report that traditional reference genes such as U6 spliceosomal RNA, Beta Actin and Glyceraldehyde-3P-dehydrogenase altered their expression in response to classic hepatotoxins and therefore cannot be used as reference genes in hepatotoxicity studies. Stability rankings of candidate reference genes, considering only those that did not alter their expression, were determined using geNorm, NormFinder and BestKeeper software packages. The potential candidates whose measurements were stable were further tested in different combinations to find the optimal set of reference genes that accurately determine mRNA and miRNA levels. Finally, the combination of MicroRNA-16/5S Ribosomal RNA and Beta 2 Microglobulin/18S Ribosomal RNA were validated as optimal reference genes for microRNA and mRNA quantification, respectively, in rat models of acute hepatotoxicity.