GeneSelectML: a comprehensive way of gene selection for RNA-Seq data via machine learning algorithms.

Selection of differentially expressed genes (DEGs) is a vital process to discover the causes of diseases. It has been shown that modelling of genomics data by considering relation among genes increases the predictive performance of methods compared to univariate analysis. However, there exist serious differences among most studies analyzing the same dataset for the reasons arising from the methods. Therefore, there is a strong need for easily accessible, user-friendly, and interactive tool to perform gene selection for RNA-seq data via machine learning algorithms simultaneously not to miss DEGs. We develop an open-source and freely available web-based tool for gene selection via machine learning algorithms that can deal with high performance computation. This tool includes six machine learning algorithms having different aspects. Moreover, the tool involves classical pre-processing steps; filtering, normalization, transformation, and univariate analysis. It also offers well-arranged graphical approaches; network plot, heatmap, venn diagram, and box-and-whisker plot. Gene ontology analysis is provided for both mRNA and miRNA DEGs. The implementation is carried out on Alzheimer RNA-seq data to demonstrate the use of this web-based tool. Eleven genes are suggested by at least two out of six methods. One of these genes, hsa-miR-148a-3p, might be considered as a new biomarker for Alzheimer's disease diagnosis. Kidney Chromophobe dataset is also analyzed to demonstrate the validity of GeneSelectML web tool on a different dataset. GeneSelectML is distinguished in that it simultaneously uses different machine learning algorithms for gene selection and can perform pre-processing, graphical representation, and gene ontology analyses on the same tool. This tool is freely available at www.softmed.hacettepe.edu.tr/GeneSelectML .

Prevalence of oral abnormalities in a Turkish newborn population.

OBJECTIVE: There are a few studies that describe the oral findings in newborn children in various populations but none conducted for a Turkish population. Hence, this study determined the prevalence of intraoral findings in a group of newborns and examined the correlation among these findings with the mother's systemic and gestational medical complications, cigarette consumption during pregnancy and consanguinity between the parents. METHODS: 2,021 full-term, newborn children were examined. Oral cysts, ankyloglossia, attached upper midline frenum and other medical diagnoses at birth were investigated. Medical information for each child and parent was recorded via standard questionnaire. Obtained data was analysed using the Pearson Chi-Square test (P≤0.05). RESULTS: The most common findings were of oral inclusion cysts situated palatally. CONCLUSIONS: There was a statistically significant relationship between the presence of oral inclusion cysts with the congenital diabetes and also insulin treatment and cigarette consumption during pregnancy. Moreover, a significant relationship was found between the presence of oral inclusion cysts and gestational diabetes and with the presence of consanguinity between the parents (P=0.004).

Predicting intraoperative pain in emergency endodontic patients: clinical study.

This prospective observational study sought to investigate the incidence of intraoperative pain (IOP) among emergency endodontic patients and to construct an IOP prediction model that includes preoperative pain level (PPL). All patients who underwent emergency endodontic treatment at Gazi University, Ankara, Turkey, during the spring term of 2016 were considered for inclusion in the study. Demographic and clinical variables and PPL were recorded. Local anesthesia was provided to all patients before beginning routine endodontic treatment. IOP was defined as the condition of requiring supplementary anesthesia before the working length was established and exhibiting persistent moderate or severe pain despite supplementary anesthesia. Data from 85% and 15% of 435 patients (178 men, 257 women; mean age: 35 years) were used to develop predictive models by multiple logistic regression analysis and to test external validity of the models, respectively. Two multiple logistic regression models achieved good model fits. Model 1 included age, pulpal diagnosis, and arc (p < 0.05). In addition to these variables, Model 2 included periapical diagnosis and PPL (p < 0.15). Models 1 and 2 showed accuracies of 0.76 and 0.75, sensitivities of 0.74 and 0.77, and specificities of 0.76 and 0.74, respectively for the modeling data (internal validity), and accuracies of 0.82 and 0.80, sensitivities of 0.83 and 0.67, and specificities of 0.81 and 0.81, respectively for the control data (external validity). The IOP incidence was 10.3%. IOP in patients undergoing emergency endodontic treatment can be successfully predicted by using models that account for demographic and clinical variables, including PPL.

Predictive Model of Intraoperative Pain during Endodontic Treatment: Prospective Observational Clinical Study.

INTRODUCTION: This observational study sought to assess the incidence of intraoperative pain (IOP) among patients receiving endodontic treatment and to construct a model for predicting the probability of IOP. METHODS: All patients attending the endodontic training clinic at Gazi University, Ankara, Turkey, during the spring term of 2014 were examined (N = 2785 patients; observation completed in 1435 patients; male: 628, female: 807; mean age: 39 years; 1655 teeth total). Demographic and clinical variables were recorded for patients requiring primary endodontic treatment. Local anesthesia was administered and routine endodontic treatment commenced. After the working length was established, each patient was asked to report any pain according to a visual analog scale. Supplementary local infiltration anesthesia was administered if necessary. If pain continued despite supplementary anesthesia, then the pain score was immediately assessed. A visual analog scale score corresponding to more than mild pain indicated IOP. A predictive model was constructed with multiple logistic regression analysis from the data of 85% of cases, with the remaining 15% of cases being used to test the external validity of the model. RESULTS: The incidence of IOP was 6.1% (101/1655 cases). One tooth from each patient was randomly selected, with 1435 teeth being retained for further analysis. A multiple logistic regression model was constructed with the variables age, tooth type, arc, pulpal diagnosis, pain present within the previous 24 hours, and anesthetic solution (P < .05). Good fits were obtained for the final model and external control, with a correct classification rate (efficiency) of 0.78, sensitivity (true positive rate) of 0.63, and specificity (true negative rate) of 0.79 for the external control. CONCLUSIONS: A successful predictive model of IOP was constructed with demographic and clinical variables.

From RNA isolation to microarray analysis: Comparison of methods in FFPE tissues.

BACKGROUND: Genome-wide gene expression profiling analysis of FFPE tissue samples is indispensable for cancer research and provides the opportunity to evaluate links between molecular and clinical information, however, working with FFPE samples is challenging due to extensive cross-linking, fragmentation and limited quantities of nucleic acid. Thus, processing of FFPE tissue samples from RNA extraction to microarray analysis still needs optimization. MATERIALS AND METHODS: In this study, a modified deparaffinization protocol was conducted prior to RNA isolation. Trizol, Qiagen RNeasy FFPE and Arcturus PicoPure RNA Isolation kits were used in parallel to compare their impact on RNA isolation. We also evaluated the effect of two different cRNA/cDNA preparation and labeling protocols with two different array platforms (Affymetrix Human Genome U133 Plus 2.0 and U133_X3P) on the percentage of present calls. RESULTS: Our optimization study shows that the Qiagen RNeasy FFPE kit with modified deparaffinization step gives better results (RNA quantity and quality) than the other two isolation kits. The Ribo-SPIA protocol gave a significantly higher percentage of present calls than the 3' IVT cDNA amplification and labeling system. However, no significant differences were found between the two array platforms. CONCLUSION: Our study paves the way for future high-throughput transcriptional analysis by optimizing FFPE tissue sample processing from RNA isolation to microarray analysis.

Flexible multivariate marginal models for analyzing multivariate longitudinal data, with applications in R.

Most of the available multivariate statistical models dictate on fitting different parameters for the covariate effects on each multiple responses. This might be unnecessary and inefficient for some cases. In this article, we propose a modelling framework for multivariate marginal models to analyze multivariate longitudinal data which provides flexible model building strategies. We show that the model handles several response families such as binomial, count and continuous. We illustrate the model on the Kenya Morbidity data set. A simulation study is conducted to examine the parameter estimates. An R package mmm2 is proposed to fit the model.

mmm: an R package for analyzing multivariate longitudinal data with multivariate marginal models.

Modeling multivariate longitudinal data has many challenges in terms of both statistical and computational aspects. Statistical challenges occur due to complex dependence structures. Computational challenges are due to the complex algorithms, the use of numerical methods, and potential convergence problems. Therefore, there is a lack of software for such data. This paper introduces an R package mmm prepared for marginal modeling of multivariate longitudinal data. Parameter estimations are achieved by generalized estimating equations approach. A real life data set is applied to illustrate the core features of the package, and sample R code snippets are provided. It is shown that the multivariate marginal models considered in this paper and mmm are valid for binary, continuous and count multivariate longitudinal responses.

Power analysis of C-TDT for small sample size genome-wide association studies by the joint use of case-parent trios and pairs.

In family-based genetic association studies, it is possible to encounter missing genotype information for one of the parents. This leads to a study consisting of both case-parent trios and case-parent pairs. One of the approaches to this problem is permutation-based combined transmission disequilibrium test statistic. However, it is still unknown how powerful this test statistic is with small sample sizes. In this paper, a simulation study is carried out to estimate the power and false positive rate of this test across different sample sizes for a family-based genome-wide association study. It is observed that a statistical power of over 80% and a reasonable false positive rate estimate can be achieved even with a combination of 50 trios and 30 pairs when 2% of the SNPs are assumed to be associated. Moreover, even smaller samples provide high power when smaller percentages of SNPs are associated with the disease.

Predicting intraoperative pain in emergency endodontic patients: clinical study

Abstract This prospective observational study sought to investigate the incidence of intraoperative pain (IOP) among emergency endodontic patients and to construct an IOP prediction model that includes preoperative pain level (PPL). All patients who underwent emergency endodontic treatment at Gazi University, Ankara, Turkey, during the spring term of 2016 were considered for inclusion in the study. Demographic and clinical variables and PPL were recorded. Local anesthesia was provided to all patients before beginning routine endodontic treatment. IOP was defined as the condition of requiring supplementary anesthesia before the working length was established and exhibiting persistent moderate or severe pain despite supplementary anesthesia. Data from 85% and 15% of 435 patients (178 men, 257 women; mean age: 35 years) were used to develop predictive models by multiple logistic regression analysis and to test external validity of the models, respectively. Two multiple logistic regression models achieved good model fits. Model 1 included age, pulpal diagnosis, and arc (p < 0.05). In addition to these variables, Model 2 included periapical diagnosis and PPL (p < 0.15). Models 1 and 2 showed accuracies of 0.76 and 0.75, sensitivities of 0.74 and 0.77, and specificities of 0.76 and 0.74, respectively for the modeling data (internal validity), and accuracies of 0.82 and 0.80, sensitivities of 0.83 and 0.67, and specificities of 0.81 and 0.81, respectively for the control data (external validity). The IOP incidence was 10.3%. IOP in patients undergoing emergency endodontic treatment can be successfully predicted by using models that account for demographic and clinical variables, including PPL.

Genome-wide transcriptional reorganization associated with senescence-to-immortality switch during human hepatocellular carcinogenesis.

Senescence is a permanent proliferation arrest in response to cell stress such as DNA damage. It contributes strongly to tissue aging and serves as a major barrier against tumor development. Most tumor cells are believed to bypass the senescence barrier (become "immortal") by inactivating growth control genes such as TP53 and CDKN2A. They also reactivate telomerase reverse transcriptase. Senescence-to-immortality transition is accompanied by major phenotypic and biochemical changes mediated by genome-wide transcriptional modifications. This appears to happen during hepatocellular carcinoma (HCC) development in patients with liver cirrhosis, however, the accompanying transcriptional changes are virtually unknown. We investigated genome-wide transcriptional changes related to the senescence-to-immortality switch during hepatocellular carcinogenesis. Initially, we performed transcriptome analysis of senescent and immortal clones of Huh7 HCC cell line, and identified genes with significant differential expression to establish a senescence-related gene list. Through the analysis of senescence-related gene expression in different liver tissues we showed that cirrhosis and HCC display expression patterns compatible with senescent and immortal phenotypes, respectively; dysplasia being a transitional state. Gene set enrichment analysis revealed that cirrhosis/senescence-associated genes were preferentially expressed in non-tumor tissues, less malignant tumors, and differentiated or senescent cells. In contrast, HCC/immortality genes were up-regulated in tumor tissues, or more malignant tumors and progenitor cells. In HCC tumors and immortal cells genes involved in DNA repair, cell cycle, telomere extension and branched chain amino acid metabolism were up-regulated, whereas genes involved in cell signaling, as well as in drug, lipid, retinoid and glycolytic metabolism were down-regulated. Based on these distinctive gene expression features we developed a 15-gene hepatocellular immortality signature test that discriminated HCC from cirrhosis with high accuracy. Our findings demonstrate that senescence bypass plays a central role in hepatocellular carcinogenesis engendering systematic changes in the transcription of genes regulating DNA repair, proliferation, differentiation and metabolism.

A novel approach for small sample size family-based association studies: sequential tests.

In this paper, we propose a sequential probability ratio test (SPRT) to overcome the problem of limited samples in studies related to complex genetic diseases. The results of this novel approach are compared with the ones obtained from the traditional transmission disequilibrium test (TDT) on simulated data. Although TDT classifies single-nucleotide polymorphisms (SNPs) to only two groups (SNPs associated with the disease and the others), SPRT has the flexibility of assigning SNPs to a third group, that is, those for which we do not have enough evidence and should keep sampling. It is shown that SPRT results in smaller ratios of false positives and negatives, as well as better accuracy and sensitivity values for classifying SNPs when compared with TDT. By using SPRT, data with small sample size become usable for an accurate association analysis.

Adhesion to primary and permanent dentin and a simple model approach.

OBJECTIVES: First to compare different dentin bonding agents' shear bond strength to primary and permanent dentin. Secondly to compare the fracture failure modes and making an attempt to develop a statistical model that could be helpful in predicting them. METHODS: Extracted human primary and permanent molars were used as substrates (dentin). The shear bond strength of composite to substrate was measured and fracture surfaces were evaluated visually and with stereomicroscope. Using the data obtained, a statistical model was built in order to predict the failure modes. RESULTS: Higher bond strength values were obtained for permanent dentin. Total-etch adhesives displayed higher shear bond strength values than the self-etch adhesive. Adhesive failures were more frequently seen in primary dentin. Self-etch adhesive system displayed more adhesive failures. Prepared model confirmed the negative relationship between shear bond strength and the probability of observing adhesive failure. CONCLUSIONS: There should be an application protocol for the usage of dentin bonding agents in primary dentin. Further development of statistical and fuzzy models for failure modes can be supportive alternatives for microscopic evaluations and also be helpful in understanding and eliminating the factors which are responsible for the formation of adhesive failures and for achieving clinically more successful adhesive restorations.