Training Test Set Research Articles

A class of joint models for multivariate longitudinal measurements and a binary event.

Predicting binary events such as newborns with large birthweight is important for obstetricians in their attempt to reduce both maternal and fetal morbidity and mortality. Such predictions have been a challenge in obstetric practice, where longitudinal ultrasound measurements taken at multiple gestational times during pregnancy may be useful for predicting various poor pregnancy outcomes. The focus of this article is on developing a flexible class of joint models for the multivariate longitudinal ultrasound measurements that can be used for predicting a binary event at birth. A skewed multivariate random effects model is proposed for the ultrasound measurements, and the skewed generalized t-link is assumed for the link function relating the binary event and the underlying longitudinal processes. We consider a shared random effect to link the two processes together. Markov chain Monte Carlo sampling is used to carry out Bayesian posterior computation. Several variations of the proposed model are considered and compared via the deviance information criterion, the logarithm of pseudomarginal likelihood, and with a training-test set prediction paradigm. The proposed methodology is illustrated with data from the NICHD Successive Small-for-Gestational-Age Births study, a large prospective fetal growth cohort conducted in Norway and Sweden.

An Open-Ended Approach to BCI: Embracing Individual Differences by Allowing for User-Defined Mental Commands

An Open-Ended Approach to BCI: Embracing Individual Differences by Allowing for User-Defined Mental Commands

A Computer Vision System for the Automatic Identification of Butterfly Species via Gabor-Filter-Based Texture Features and Extreme Learning Machine: GF+ELM

Butterflies are classified first according to their outer morphological qualities. It is required to analyze their genital characters when classification according to their outer morphological qualities is not possible. The genital characters of butterflies can be obtained using various chemical substances and methods; however, these processes can only be carried out with some certain expenses. Furthermore, the preparation of genital slides is time-consuming since it requires specific processes. In this study, a new method based on the extreme learning machine (ELM) and Gabor filters (GFs), which is an image processing technique, was used for the identification of butterfly species as an alternative to conventional diagnostic methods. GFs have been recognized as a very useful tool in texture analysis, due to their optimal localization properties in both the spatial and frequency domains, and have been found to have widespread use in computer vision applications. To obtain the appropriate features from butterfly images in the spatial domain, 20 filters were designed for the various angles and frequencies (5 frequencies and 4 orientations). The diagnosing of butterflies was performed through ELM, with texture features based on GFs. The classification process was performed with a 75%–25% training-test set for different activation functions and the recognition performance value was obtained as 97.00%. In addition, the recognition success rates with ELM were compared to other machine learning methods and it was seen that ELM has a more significant success rate in butterfly identification than other methods. As a result, the proposed method is a suitable machine vision system for detecting butterfly species

Multi-conformation 3D QSAR study of benzenesulfonyl-pyrazol-ester compounds and their analogs as cathepsin B inhibitors

Cathepsin B has been found being responsible for many human diseases. Inhibitors of cathepsin B, a ubiquitous lysosomal cysteine protease, have been developed as a promising treatment for human diseases resulting from malfunction and over-expression of this enzyme. Through a high throughput screening assay, a set of compounds were found able to inhibit the enzymatic activity of cathepsin B. The binding structures of these active compounds were modeled through docking simulation. Three-dimensional (3D) quantitative structure–activity relationship (QSAR) models were constructed using comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) based on the docked structures of the compounds. Strong correlations were obtained for both CoMFA and CoMSIA models with cross-validated correlation coefficients ( q 2) of 0.605 and 0.605 and the regression correlation coefficients ( r 2) of 0.999 and 0.997, respectively. The robustness of these models was further validated using leave-one-out (LOO) method and training-test set method. The activities of eight (8) randomly selected compounds were predicted using models built from training set of compounds with prediction errors of less than 1 unit for most compounds in CoMFA and CoMSIA models. Structural features for compounds with improved activity are suggested based on the analysis of the CoMFA and CoMSIA contour maps and the property map of the protein ligand binding site. These results may help to provide better understanding of the structure–activity relationship of cathepsin B inhibitors and to facilitate lead optimization and novel inhibitor design. The multi-conformation method to build 3D QSAR is very effective approach to obtain satisfactory models with high correlation with experimental results and high prediction power for unknown compounds.

Simple strategies for haplotype analysis of the X chromosome with application to age-related macular degeneration

For haplotype analysis of the X chromosome, haplotype-sharing (HS) statistics with sliding windows are defined for males and females separately, which are then combined to a single HS test for the X chromosome. When independent replication samples are not available, the training-testing sets approach is used to validate this procedure and a permutation method is used to obtain its P-value. We applied this method to the X chromosome (with 1804 SNPs) for age-related macular degeneration (AMD). We found a window of five SNPs over a 272 kb region associated with AMD after Bonferroni correction. An examination of the odds ratio and the population attributable risks revealed a disease-preventive haplotype, ATGAC, on these five SNPs. For elderly females without this haplotype, the likelihood of AMD is increased by a factor of 4.75 with a 95% confidence interval (1.43, 15.82). The frequency of ATGAC in HapMap CEU is 0.276. These five SNPs are covered by the gene DIAPH2, which is known to cause premature ovarian failure (POF) in females. Our results indicated that DIAPH2 may be a polygenic pleiotropy for POF and AMD.

QSAR models for predicting cathepsin B inhibition by small molecules—Continuous and binary QSAR models to classify cathepsin B inhibition activities of small molecules

Cathepsin B is a potential target for the development of drugs to treat several important human diseases. A number of inhibitors targeting this protein have been developed in the past several years. Recently, a group of small molecules were identified to have inhibitory activity against cathepsin B through high throughput screening (HTS) tests. In this study, traditional continuous and binary QSAR models were built to classify the biological activities of previously identified compounds and to distinguish active compounds from inactive compounds for drug development based on the calculated molecular and physicochemical properties. Strong correlations were obtained for the continuous QSAR models with regression correlation coefficients ( r 2) and cross-validated correlation coefficients ( q 2) of 0.77 and 0.61 for all compounds, and 0.82 and 0.68 for the compound set excluding 3 outliers, respectively. The models were further validated through the leave-one-out (LOO) method and the training-test set method. The binary models demonstrated a strong level of predictability in distinguishing the active compounds from inactive compounds with accuracies of 0.89 and 0.94 for active and inactive compounds, respectively, in non-cross-validated models. Similar results were obtained for the cross-validated models. Collectively, these results demonstrate the models’ ability to discriminate between active and inactive compounds, suggesting that the models may be used to pre-screen compounds to facilitate compound optimization and to design novel inhibitors for drug development.

Derivation of Stable Microarray Cancer-Differentiating Signatures Using Consensus Scoring of Multiple Random Sampling and Gene-Ranking Consistency Evaluation

Microarrays have been explored for deriving molecular signatures to determine disease outcomes, mechanisms, targets, and treatment strategies. Although exhibiting good predictive performance, some derived signatures are unstable due to noises arising from measurement variability and biological differences. Improvements in measurement, annotation, and signature selection methods have been proposed. We explored a new signature selection method that incorporates consensus scoring of multiple random sampling and multistep evaluation of gene-ranking consistency for maximally avoiding erroneous elimination of predictor genes. This method was tested by using a well-studied 62-sample colon cancer data set and two other cancer data sets (86-sample lung adenocarcinoma and 60-sample hepatocellular carcinoma). For the colon cancer data set, the derived signatures of 20 sampling sets, composed of 10,000 training test sets, are fairly stable with 80% of top 50 and 69% to 93% of all predictor genes shared by all 20 signatures. These shared predictor genes include 48 cancer-related and 16 cancer-implicated genes, as well as 50% of the previously derived predictor genes. The derived signatures outperform all previously derived signatures in predicting colon cancer outcomes from an independent data set collected from the Stanford Microarray Database. Our method showed similar performance for the other two data sets, suggesting its usefulness in deriving stable signatures for biomarker and target discovery.

Prediction of a key role of motifs binding E2F and NR2F in down-regulation of numerous genes during the development of the mouse hippocampus

BackgroundWe previously demonstrated that gene expression profiles during neuronal differentiation in vitro and hippocampal development in vivo were very similar, due to a conservation of the important second singular value decomposition (SVD) mode (Mode 2) of expression. The conservation of Mode 2 suggests that it reflects a regulatory mechanism conserved between the two systems. In either dataset, the expression vectors of all the genes form two large clusters that differ in the sign of the contribution of Mode 2, which for the majority of them reflects the difference between down- or up-regulation.ResultsIn the current work, we used a novel approach of analyzing cis-regulation of gene expression in a subspace of a single SVD mode of temporal expression profiles. In the putative upstream regulatory sequences identified by mouse-human homology for all the genes represented in either dataset, we searched for simple features (motifs and pairs of motifs) associated with either sign of the loading of Mode 2. Using a cross-system training-test set approach, we identified E2F binding sites as predictors of down-regulation of gene expression during hippocampal development. NR2F binding sites, for the transcription factors Nr2f/COUP and Hnf4, and also NR2F_SP1 pairs of binding sites, were predictors of down-regulation of expression both during hippocampal development and neuronal differentiation. Analysis of another dataset, from gene profiling of myoblast differentiation in vitro, shows that the conservation of Mode 2 extends to the differentiation of mesenchymal cells. This permitted the identification of two more pairs of motifs, one of which included the CDE/CHR tandem element, as features associated with down-regulation both in the differentiating myoblasts and in the developing hippocampus. Of the features we identified, the E2F and CDE/CHR motifs may be associated with the cycling progenitor cell status, while NR2F may be related to the entry into differentiation along the neuronal pathway.ConclusionOur results constitute the first prediction of an expression pattern from the genomic sequence for the developing mammalian brain, and demonstrate a potential for the analysis of gene regulation in a subspace of a single SVD mode of expression.

Correlations between Injection Molding Parameters, Morphology and Mechanical Properties of PPS Using Artificial Neural Networks

Abstract The processing conditions of injection molding have a complex influence on the morphology and on the mechanical properties of a semicrystalline polymer. Therefore, to establish correlations between processing conditions, morphology and mechanical properties constitutes a difficult task; finding these correlations is one of the goals of a materials engineer. The purpose of this work was to study the influence of the injection molding conditions on the morphology and mechanical properties of poly(p-phenylene sulphide), PPS, using artificial neural networks, ANNs. First, a statistical analysis was done to find the more influential processing parameters that affect the morphology and mechanical properties of the PPS. Second, ANNs were applied to establish correlations between processing conditions, morphology and properties. It was found that the variables with the highest influence on the morphology and the mechanical properties were the injection and mold temperatures (Tinj and Tmold, respectively), as they showed a straight relationship with the crystallinity index of the injection molded part. Three different ANNs were built to predict the correlations. The ANN-1 predicted the crystallinity gradient along the thickness of the injection molded part from Tmold, Tinj, and flow rate, Q; the ANN-2 predicted the elastic and flexural modulus, E, and the yield stress from the crystallinity gradient, while the ANN-3 predicted the mechanical properties directly from the processing conditions. All ANNs were built with only fifteen experimental data and were trained with the group cross-validation method, GCV and with a training-test set method. Both methods showed similar and excellent performance. Thus, it can be concluded that ANNs can be used as a powerful tool in the learning of these complex correlations.