Epigenome-wide association study of peripheral blood mononuclear cells in systemic lupus erythematosus: Identifying DNA methylation signatures associated with interferon-related genes based on ethnicity and SLEDAI.

Systemic lupus erythematosus (SLE or lupus) is a heterogeneous autoimmune disease characterized by the involvement of multiple organs and the production of antinuclear antibodies. DNA methylation plays an important role in the pathogenesis of lupus. We have performed an epigenome-wide DNA methylation study in lupus and healthy control (non-lupus) subjects to identify epigenetic patterns in lupus characterized ethnicity and SLE disease activity index (SLEDAI). A total of fifty-seven lupus patients (39 African American (AA) and 18 European American (EA)) and 33 healthy controls (17 AA and 16 EA) were studied. Differential DNA methylation between lupus patients and controls was assessed for approximately 485,000 CpG sites across the genome. We identified 41 differentially methylated sites (associated with 30 genes) between lupus and control s subjects, 85% of which were hypomethylated. Significant hypomethylation of differentially methylated sites was associated with several interferon-related genes, including MX1, IFI44L, PARP9, DT3XL, IFIT1, IFI44, RSAD2, PLSCR1, and IRF7. Several of these associated genes were also hypomethylated in comparisons between AA lupus and AA non-lupus subjects and between lupus patients with SLEDAI>6 and non-lupus subjects. Our analysis of gene expression data through RT-PCR confirmed these findings. Thus, the results indicate epigenetics susceptibility in lupus, which may be associated with SLEDAI score and ethnicity. In addition, our findings support the importance of the Type 1 interferon pathway in lupus pathogenesis.

Electron microscopy techniques employed to explore mitochondrial defects in the developing rat brain following ketamine treatment.

Ketamine, an FDA-approved N-methyl-D-aspartate (NMDA) receptor antagonist, is commonly used for general pediatric anesthesia. Accumulating evidence has indicated that prolonged exposure to ketamine induces widespread apoptotic cell death in the developing brains of experimental animals. Although mitochondria are known to play a pivotal role in cell death, little is known about the alterations in mitochondrial ultrastructure that occur during ketamine-induced neurotoxicity. The objective of this pilot study was to utilize classic and contemporary methods in electron microscopy to study the impact of ketamine on the structure of mitochondria in the developing rat brain. While transmission electron microscopy (TEM) was employed to comprehensively study mitochondrial inner membrane topology, serial block-face scanning electron microscopy (SBF-SEM) was used as a complementary technique to compare the overall mitochondrial morphology from a representative treated and untreated neuron. In this study, postnatal day 7 (PND-7) Sprague-Dawley rats were treated with ketamine or saline (6 subcutaneous injections × 20 mg/kg or 10 ml/kg, respectively, at 2-h intervals with a 6-h withdrawal period after the last injection, n=6 each group). Samples from the frontal cortex were harvested and analyzed using TEM or SBF-SEM. While classic TEM revealed that repeated ketamine exposure induces significant mitochondrial swelling in neurons, the newer technique of SBF-SEM confirmed the mitochondrial swelling in three dimensions (3D) and showed that ketamine exposure may also induce mitochondrial fission, which was not observable in the two dimensions (2D) of TEM. Furthermore, 3D statistical analysis of these reconstructed mitochondria appeared to show that ketamine-treated mitochondria had significantly larger volumes per unit surface area than mitochondria from the untreated neuron. The ultrastructural mitochondrial alterations demonstrated here by TEM and SBF-SEM support ketamine's proposed mechanism of neurotoxicity in the developing rat brain.

Low-Frequency Mutational Heterogeneity of Invasive Ductal Carcinoma Subtypes: Information to Direct Precision Oncology.

Information regarding the role of low-frequency hotspot cancer-driver mutations (CDMs) in breast carcinogenesis and therapeutic response is limited. Using the sensitive and quantitative Allele-specific Competitor Blocker PCR (ACB-PCR) approach, mutant fractions (MFs) of six CDMs (PIK3CA H1047R and E545K, KRAS G12D and G12V, HRAS G12D, and BRAF V600E) were quantified in invasive ductal carcinomas (IDCs; including ~20 samples per subtype). Measurable levels (i.e., ≥ 1 × 10-5, the lowest ACB-PCR standard employed) of the PIK3CA H1047R, PIK3CA E545K, KRAS G12D, KRAS G12V, HRAS G12D, and BRAF V600E mutations were observed in 34/81 (42%), 29/81 (36%), 51/81 (63%), 9/81 (11%), 70/81 (86%), and 48/81 (59%) of IDCs, respectively. Correlation analysis using available clinicopathological information revealed that PIK3CA H1047R and BRAF V600E MFs correlate positively with maximum tumor dimension. Analysis of IDC subtypes revealed minor mutant subpopulations of critical genes in the MAP kinase pathway (KRAS, HRAS, and BRAF) were prevalent across IDC subtypes. Few triple-negative breast cancers (TNBCs) had appreciable levels of PIK3CA mutation, suggesting that individuals with TNBC may be less responsive to inhibitors of the PI3K/AKT/mTOR pathway. These results suggest that low-frequency hotspot CDMs contribute significantly to the intertumoral and intratumoral genetic heterogeneity of IDCs, which has the potential to impact precision oncology approaches.

Corticosterone and exogenous glucose alter blood glucose levels, neurotoxicity, and vascular toxicity produced by methamphetamine.

Our previous studies have raised the possibility that altered blood glucose levels may influence and/or be predictive of methamphetamine (METH) neurotoxicity. This study evaluated the effects of exogenous glucose and corticosterone (CORT) pretreatment alone or in combination with METH on blood glucose levels and the neural and vascular toxicity produced. METH exposure consisted of four sequential injections of 5, 7.5, 10, and 10 mg/kg (2 h between injections) D-METH. The three groups given METH in combination with saline, glucose (METH+Glucose), or CORT (METH+CORT) had significantly higher glucose levels compared to the corresponding treatment groups without METH except at 3 h after the last injection. At this last time point, the METH and METH+Glucose groups had lower levels than the non-METH groups, while the METH+CORT group did not. CORT alone or glucose alone did not significantly increase blood glucose. Mortality rates for the METH+CORT (40%) and METH+Glucose (44%) groups were substantially higher than the METH (< 10%) group. Additionally, METH+CORT significantly increased neurodegeneration above the other three METH treatment groups (≈ 2.5-fold in the parietal cortex). Thus, maintaining elevated levels of glucose during METH exposure increases lethality and may exacerbate neurodegeneration. Neuroinflammation, specifically microglial activation, was associated with degenerating neurons in the parietal cortex and thalamus after METH exposure. The activated microglia in the parietal cortex were surrounding vasculature in most cases and the extent of microglial activation was exacerbated by CORT pretreatment. Our findings show that acute CORT exposure and elevated blood glucose levels can exacerbate METH-induced vascular damage, neuroinflammation, neurodegeneration and lethality. Cover Image for this issue: doi. 10.1111/jnc.13819.

Early metabolomics changes in heart and plasma during chronic doxorubicin treatment in B6C3F1 mice.

The present study aimed to identify molecular markers of early stages of cardiotoxicity induced by a potent chemotherapeutic agent, doxorubicin (DOX). Male B6C3F1 mice were dosed with 3 mg kg(-1) DOX or saline via tail vein weekly for 2, 3, 4, 6 or 8 weeks (cumulative DOX doses of 6, 9, 12, 18 or 24 mg kg(-1) , respectively) and euthanized a week after the last dose. Mass spectrometry-based and nuclear magnetic resonance spectrometry-based metabolic profiling were employed to identify initial biomarkers of cardiotoxicity before myocardial injury and cardiac pathology, which were not noted until after the 18 and 24 mg kg(-1) cumulative doses, respectively. After a cumulative dose of 6 mg kg(-1) , 18 amino acids and four biogenic amines (acetylornithine, kynurenine, putrescine and serotonin) were significantly increased in cardiac tissue; 16 amino acids and two biogenic amines (acetylornithine and hydroxyproline) were significantly altered in plasma. In addition, 16 acylcarnitines were significantly increased in plasma and five were significantly decreased in cardiac tissue compared to saline-treated controls. Plasma lactate and succinate, involved in the Krebs cycle, were significantly altered after a cumulative dose of 6 mg kg(-1) . A few metabolites remained altered at higher cumulative DOX doses, which could partly indicate a transition from injury processes at 2 weeks to repair processes with additional injury happening concurrently before myocardial injury at 8 weeks. These altered metabolic profiles in mouse heart and plasma during the initial stages of injury progression due to DOX treatment may suggest these metabolites as candidate early biomarkers of cardiotoxicity. Published 2016. This article is a U.S. Government work and is in the public domain in the USA.

DAFS: a data-adaptive flag method for RNA-sequencing data to differentiate genes with low and high expression.

BACKGROUND: Next-generation sequencing (NGS) has advanced the application of high-throughput sequencing technologies in genetic and genomic variation analysis. Due to the large dynamic range of expression levels, RNA-seq is more prone to detect transcripts with low expression. It is clear that genes with no mapped reads are not expressed; however, there is ongoing debate about the level of abundance that constitutes biologically meaningful expression. To date, there is no consensus on the definition of low expression. Since random variation is high in regions with low expression and distributions of transcript expression are affected by numerous experimental factors, methods to differentiate low and high expressed data in a sample are critical to interpreting classes of abundance levels in RNA-seq data. RESULTS: A data-adaptive approach was developed to estimate the lower bound of high expression for RNA-seq data. The Kolmgorov-Smirnov statistic and multivariate adaptive regression splines were used to determine the optimal cutoff value for separating transcripts with high and low expression. Results from the proposed method were compared to results obtained by estimating the theoretical cutoff of a fitted two-component mixture distribution. The robustness of the proposed method was demonstrated by analyzing different RNA-seq datasets that varied by sequencing depth, species, scale of measurement, and empirical density shape. CONCLUSIONS: The analysis of real and simulated data presented here illustrates the need to employ data-adaptive methodology in lieu of arbitrary cutoffs to distinguish low expressed RNA-seq data from high expression. Our results also present the drawbacks of characterizing the data by a two-component mixture distribution when classes of gene expression are not well separated. The ability to ascertain stably expressed RNA-seq data is essential in the filtering process of data analysis, and methodologies that consider the underlying data structure demonstrate superior performance in preserving most of the interpretable and meaningful data. The proposed algorithm for classifying low and high regions of transcript abundance promises wide-range application in the continuing development of RNA-seq analysis.

Comparison of the global gene expression of choroid plexus and meninges and associated vasculature under control conditions and after pronounced hyperthermia or amphetamine toxicity.

BACKGROUND: The meninges (arachnoid and pial membranes) and associated vasculature (MAV) and choroid plexus are important in maintaining cerebrospinal fluid (CSF) generation and flow. MAV vasculature was previously observed to be adversely affected by environmentally-induced hyperthermia (EIH) and more so by a neurotoxic amphetamine (AMPH) exposure. Herein, microarray and RT-PCR analysis was used to compare the gene expression profiles between choroid plexus and MAV under control conditions and at 3 hours and 1 day after EIH or AMPH exposure. Since AMPH and EIH are so disruptive to vasculature, genes related to vasculature integrity and function were of interest. RESULTS: Our data shows that, under control conditions, many of the genes with relatively high expression in both the MAV and choroid plexus are also abundant in many epithelial tissues. These genes function in transport of water, ions, and solutes, and likely play a role in CSF regulation. Most genes that help form the blood-brain barrier (BBB) and tight junctions were also highly expressed in MAV but not in choroid plexus. In MAV, exposure to EIH and more so to AMPH decreased the expression of BBB-related genes such as Sox18, Ocln, and Cldn5, but they were much less affected in the choroid plexus. There was a correlation between the genes related to reactive oxidative stress and damage that were significantly altered in the MAV and choroid plexus after either EIH or AMPH. However, AMPH (at 3 hr) significantly affected about 5 times as many genes as EIH in the MAV, while in the choroid plexus EIH affected more genes than AMPH. Several unique genes that are not specifically related to vascular damage increased to a much greater extent after AMPH compared to EIH in the MAV (Lbp, Reg3a, Reg3b, Slc15a1, Sct and Fst) and choroid plexus (Bmp4, Dio2 and Lbp). CONCLUSIONS: Our study indicates that the disruption of choroid plexus function and damage produced by AMPH and EIH is significant, but the changes may not be as pronounced as they are in the MAV, particularly for AMPH. Expression profiles in the MAV and choroid plexus differed to some extent and differences were not restricted to vascular related genes.

Development of doxorubicin-induced chronic cardiotoxicity in the B6C3F1 mouse model.

Serum levels of cardiac troponins serve as biomarkers of myocardial injury. However, troponins are released into the serum only after damage to cardiac tissue has occurred. Here, we report development of a mouse model of doxorubicin (DOX)-induced chronic cardiotoxicity to aid in the identification of predictive biomarkers of early events of cardiac tissue injury. Male B6C3F(1) mice were administered intravenous DOX at 3mg/kg body weight, or an equivalent volume of saline, once a week for 4, 6, 8, 10, 12, and 14weeks, resulting in cumulative DOX doses of 12, 18, 24, 30, 36, and 42mg/kg, respectively. Mice were sacrificed a week following the last dose. A significant reduction in body weight gain was observed in mice following exposure to a weekly DOX dose for 1week and longer compared to saline-treated controls. DOX treatment also resulted in declines in red blood cell count, hemoglobin level, and hematocrit compared to saline-treated controls after the 2nd weekly dose until the 8th and 9th doses, followed by a modest recovery. All DOX-treated mice had significant elevations in cardiac troponin T concentrations in plasma compared to saline-treated controls, indicating cardiac tissue injury. Also, a dose-related increase in the severity of cardiac lesions was seen in mice exposed to 24mg/kg DOX and higher cumulative doses. Mice treated with cumulative DOX doses of 30mg/kg and higher showed a significant decline in heart rate, suggesting drug-induced cardiac dysfunction. Altogether, these findings demonstrate the development of DOX-induced chronic cardiotoxicity in B6C3F(1) mice.

Model uncertainty and model averaging in the estimation of infectious doses for microbial pathogens.

Food-borne infection is caused by intake of foods or beverages contaminated with microbial pathogens. Dose-response modeling is used to estimate exposure levels of pathogens associated with specific risks of infection or illness. When a single dose-response model is used and confidence limits on infectious doses are calculated, only data uncertainty is captured. We propose a method to estimate the lower confidence limit on an infectious dose by including model uncertainty and separating it from data uncertainty. The infectious dose is estimated by a weighted average of effective dose estimates from a set of dose-response models via a Kullback information criterion. The confidence interval for the infectious dose is constructed by the delta method, where data uncertainty is addressed by a bootstrap method. To evaluate the actual coverage probabilities of the lower confidence limit, a Monte Carlo simulation study is conducted under sublinear, linear, and superlinear dose-response shapes that can be commonly found in real data sets. Our model-averaging method achieves coverage close to nominal in almost all cases, thus providing a useful and efficient tool for accurate calculation of lower confidence limits on infectious doses.

Physiological effects following administration of Citrus aurantium for 28 days in rats.

BACKGROUND: Since ephedra-containing dietary supplements were banned from the US market, manufacturers changed their formulations by eliminating ephedra and replacing with other botanicals, including Citrus aurantium, or bitter orange. Bitter orange contains, among other compounds, synephrine, a chemical that is chemically similar to ephedrine. Since ephedrine may have cardiovascular effects, the goal of this study was to investigate the cardiovascular effects of various doses of bitter orange extract and pure synephrine in rats. METHOD: Female Sprague-Dawley rats were dosed daily by gavage for 28 days with synephrine from two different extracts. One extract contained 6% synephrine, and the other extract contained 95% synephrine. Doses were 10 or 50mg synephrine/kg body weight from each extract. Additionally, caffeine was added to these doses, since many dietary supplements also contain caffeine. Telemetry was utilized to monitor heart rate, blood pressure, body temperature and QT interval in all rats. RESULTS AND CONCLUSION: Synephrine, either as the bitter orange extract or as pure synephrine, increased heart rate and blood pressure. Animals treated with 95% synephrine showed minimal effects on heart rate and blood pressure; more significant effects were observed with the bitter orange extract suggesting that other components in the botanical can alter these physiological parameters. The increases in heart rate and blood pressure were more pronounced when caffeine was added. None of the treatments affected uncorrected QT interval in the absence of caffeine.

Evaluation of fecal mRNA reproducibility via a marginal transformed mixture modeling approach.

BACKGROUND: Developing and evaluating new technology that enables researchers to recover gene-expression levels of colonic cells from fecal samples could be key to a non-invasive screening tool for early detection of colon cancer. The current study, to the best of our knowledge, is the first to investigate and report the reproducibility of fecal microarray data. Using the intraclass correlation coefficient (ICC) as a measure of reproducibility and the preliminary analysis of fecal and mucosal data, we assessed the reliability of mixture density estimation and the reproducibility of fecal microarray data. Using Monte Carlo-based methods, we explored whether ICC values should be modeled as a beta-mixture or transformed first and fitted with a normal-mixture. We used outcomes from bootstrapped goodness-of-fit tests to determine which approach is less sensitive toward potential violation of distributional assumptions. RESULTS: The graphical examination of both the distributions of ICC and probit-transformed ICC (PT-ICC) clearly shows that there are two components in the distributions. For ICC measurements, which are between 0 and 1, the practice in literature has been to assume that the data points are from a beta-mixture distribution. Nevertheless, in our study we show that the use of a normal-mixture modeling approach on PT-ICC could provide superior performance. CONCLUSIONS: When modeling ICC values of gene expression levels, using mixture of normals in the probit-transformed (PT) scale is less sensitive toward model mis-specification than using mixture of betas. We show that a biased conclusion could be made if we follow the traditional approach and model the two sets of ICC values using the mixture of betas directly. The problematic estimation arises from the sensitivity of beta-mixtures toward model mis-specification, particularly when there are observations in the neighborhood of the the boundary points, 0 or 1. Since beta-mixture modeling is commonly used in approximating the distribution of measurements between 0 and 1, our findings have important implications beyond the findings of the current study. By using the normal-mixture approach on PT-ICC, we observed the quality of reproducible genes in fecal array data to be comparable to those in mucosal arrays.

Endoplasmic reticulum stress responses differ in meninges and associated vasculature, striatum, and parietal cortex after a neurotoxic amphetamine exposure.

Amphetamine (AMPH) is used to treat attention deficit and hyperactivity disorders, but it can produce neurotoxicity and adverse vascular effects at high doses. The endoplasmic reticulum (ER) stress response (ERSR) entails the unfolded protein response, which helps to avoid or minimize ER dysfunction. ERSR is often associated with toxicities resulting from the accumulation of unfolded or misfolded proteins and has been associated with methamphetamine toxicity in the striatum. The present study evaluates the effect of AMPH on several ERSR elements in meninges and associated vasculature (MAV), parietal cortex, and striatum. Adult, male Sprague-Dawley rats were exposed to saline, environmentally induced hyperthermia (EIH) or four consecutive doses of AMPH that produce hyperthermia. Expression changes (mRNA and protein levels) of key ERSR-related genes in MAV, striatum, and parietal cortex at 3 h or 1 day postdosing were monitored. AMPH increased the expression of some ERSR-related genes in all tissues. Atf4 (activating transcription factor 4, an indicator of Perk pathway activation), Hspa5/Grp78 (Glucose regulated protein 78, master regulator of ERSR), Pdia4 (protein disulfide isomerase, protein-folding enzyme), and Nfkb1 (nuclear factor of kappa b, ERSR sensor) mRNA increased significantly in MAV and parietal cortex 3 h after AMPH. In striatum, Atf4 and Hspa5/Grp78 mRNA significantly increased 3 h after AMPH, but Pdia4 and Nfkb11 did not. Thus, AMPH caused a robust activation of the Perk pathway in all tissues, but significant Ire1 pathway activation occurred only after AMPH treatment in the parietal cortex and striatum. Ddit3/Chop, a downstream effector of the ERSR pathway related to the neurotoxicity, was only increased in striatum and parietal cortex. Conversely, Pdia4, an enzyme protective in the ERSR, was only increased in MAV. The overall ERSR manifestation varied significantly between MAV, striatum, and parietal cortex after a neurotoxic exposure to AMPH.

Amphetamine and environmentally induced hyperthermia differentially alter the expression of genes regulating vascular tone and angiogenesis in the meninges and associated vasculature.

An amphetamine (AMPH) regimen that does not produce a prominent blood-brain barrier breakdown was shown to significantly alter the expression of genes regulating vascular tone, immune function, and angiogenesis in vasculature associated with arachnoid and pia membranes of the forebrain. Adult-male Sprague-Dawley rats were given either saline injections during environmentally-induced hyperthermia (EIH) or four doses of AMPH with 2 h between each dose (5, 7.5, 10, and 10 mg/kg d-AMPH, s.c.) that produced hyperthermia. Rats were sacrificed either 3 h or 1 day after dosing, and total RNA and protein was isolated from the meninges, arachnoid and pia membranes, and associated vasculature (MAV) that surround the forebrain. Vip, eNos, Drd1a, and Edn1 (genes regulating vascular tone) were increased by either EIH or AMPH to varying degrees in MAV, indicating that EIH and AMPH produce differential responses to enhance vasodilatation. AMPH, and EIH to a lesser extent, elicited a significant inflammatory response at 3 h as indicated by an increased MAV expression of cytokines Il1b, Il6, Ccl-2, Cxcl1, and Cxcl2. Also, genes related to heat shock/stress and disruption of vascular homeostasis such as Icam1 and Hsp72 were also observed. The increased expression of Ctgf and Timp1 and the decreased expression of Akt1, Anpep, and Mmp2 and Tek (genes involved in stimulating angiogenesis) from AMPH exposure suggest that angiogenesis was arrested or disrupted in MAV to a greater extent by AMPH compared to EIH. Alterations in vascular-related gene expression in the parietal cortex and striatum after AMPH were less in magnitude than in MAV, indicating less of a disruption of vascular homeostasis in these two regions. Changes in the levels of insulin-like growth factor binding proteins Igfbp1, 2, and 5 in MAV, compared to those in striatum and parietal cortex, imply an interaction between these regions to regulate the levels of insulin-like growth factor after AMPH damage. Thus, the vasculature and meninges surrounding the surface of the forebrain may be an important region in which AMPHs can disrupt vascular homeostasis.

Identification of whole blood mRNA and microRNA biomarkers of tissue damage and immune function resulting from amphetamine exposure or heat stroke in adult male rats.

This work extends the understanding of how toxic exposures to amphetamine (AMPH) adversely affect the immune system and lead to tissue damage. Importantly, it determines which effects of AMPH are and are not due to pronounced hyperthermia. Whole blood messenger RNA (mRNA) and whole blood and serum microRNA (miRNA) transcripts were identified in adult male Sprague-Dawley rats after exposure to toxic AMPH under normothermic conditions, AMPH when it produces pronounced hyperthermia, or environmentally-induced hyperthermia (EIH). mRNA transcripts with large increases in fold-change in treated relative to control rats and very low expression in the control group were a rich source of organ-specific transcripts in blood. When severe hyperthermia was produced by either EIH or AMPH, significant increases in circulating organ-specific transcripts for liver (Alb, Fbg, F2), pancreas (Spink1), bronchi/lungs (F3, Cyp4b1), bone marrow (Np4, RatNP-3b), and kidney (Cesl1, Slc22a8) were observed. Liver damage was suggested also by increased miR-122 levels in the serum. Increases in muscle/heart-enriched transcripts were produced by AMPH even in the absence of hyperthermia. Expression increases in immune-related transcripts, particularly Cd14 and Vcan, indicate that AMPH can activate the innate immune system in the absence of hyperthermia. Most transcripts specific for T-cells decreased 50-70% after AMPH exposure or EIH, with the noted exception of Ccr5 and Chst12. This is probably due to T-cells leaving the circulation and down-regulation of these genes. Transcript changes specific for B-cells or B-lymphoblasts in the AMPH and EIH groups ranged widely from decreasing ≈ 40% (Cd19, Cd180) to increasing 30 to 100% (Tk1, Ahsa1) to increasing ≥500% (Stip1, Ackr3). The marked increases in Ccr2, Ccr5, Pld1, and Ackr3 produced by either AMPH or EIH observed in vivo provide further insight into the initial immune system alterations that result from methamphetamine and AMPH abuse and could modify risk for HIV and other viral infections.

Multi-class computational evolution: development, benchmark evaluation and application to RNA-Seq biomarker discovery.

BACKGROUND: A computational evolution system (CES) is a knowledge discovery engine that can identify subtle, synergistic relationships in large datasets. Pareto optimization allows CESs to balance accuracy with model complexity when evolving classifiers. Using Pareto optimization, a CES is able to identify a very small number of features while maintaining high classification accuracy. A CES can be designed for various types of data, and the user can exploit expert knowledge about the classification problem in order to improve discrimination between classes. These characteristics give CES an advantage over other classification and feature selection algorithms, particularly when the goal is to identify a small number of highly relevant, non-redundant biomarkers. Previously, CESs have been developed only for binary class datasets. In this study, we developed a multi-class CES. RESULTS: The multi-class CES was compared to three common feature selection and classification algorithms: support vector machine (SVM), random k-nearest neighbor (RKNN), and random forest (RF). The algorithms were evaluated on three distinct multi-class RNA sequencing datasets. The comparison criteria were run-time, classification accuracy, number of selected features, and stability of selected feature set (as measured by the Tanimoto distance). The performance of each algorithm was data-dependent. CES performed best on the dataset with the smallest sample size, indicating that CES has a unique advantage since the accuracy of most classification methods suffer when sample size is small. CONCLUSION: The multi-class extension of CES increases the appeal of its application to complex, multi-class datasets in order to identify important biomarkers and features.

Assessing Sex Differences in the Risk of Cardiovascular Disease and Mortality per Increment in Systolic Blood Pressure: A Systematic Review and Meta-Analysis of Follow-Up Studies in the United States.

In the United States (US), cardiovascular (CV) disease accounts for nearly 20% of national health care expenses. Since costs are expected to increase with the aging population, informative research is necessary to address the growing burden of CV disease and sex-related differences in diagnosis, treatment, and outcomes. Hypertension is a major risk factor for CV disease and mortality. To evaluate whether there are sex-related differences in the effect of systolic blood pressure (SBP) on the risk of CV disease and mortality, we performed a systematic review and meta-analysis. We conducted a comprehensive search using PubMed and Google Scholar to identify US-based studies published prior to 31 December, 2015. We identified eight publications for CV disease risk, which provided 9 female and 8 male effect size (ES) observations. We also identified twelve publications for CV mortality, which provided 10 female and 18 male ES estimates. Our meta-analysis estimated that the pooled ES for increased risk of CV disease per 10 mmHg increment in SBP was 25% for women (95% Confidence Interval (CI): 1.18, 1.32) and 15% for men (95% CI: 1.11, 1.19). The pooled increase in CV mortality per 10 mm Hg SBP increment was similar for both women and men (Women: 1.16; 95% CI: 1.10, 1.23; Men: 1.17; 95% CI: 1.12, 1.22). After adjusting for age and baseline SBP, the results demonstrated that the risk of CV disease per 10 mm Hg SBP increment for women was 1.1-fold higher than men (P<0.01; 95% CI: 1.04, 1.17). Heterogeneity was moderate but significant. There was no significant sex difference in CV mortality.

Cost-sensitive Performance Metric for Comparing Multiple Ordinal Classifiers.

The surge of interest in personalized and precision medicine during recent years has increased the application of ordinal classification problems in biomedical science. Currently, accuracy, Kendall's τb , and average mean absolute error are three commonly used metrics for evaluating the effectiveness of an ordinal classifier. Although there are benefits to each, no single metric considers the benefits of predictive accuracy with the tradeoffs of misclassification cost. In addition, decision analysis that considers pairwise analysis of the metrics is not trivial due to inconsistent findings. A new cost-sensitive metric is proposed to find the optimal tradeoff between the two most critical performance measures of a classification task - accuracy and cost. The proposed method accounts for an inherent ordinal data structure, total misclassification cost of a classifier, and imbalanced class distribution. The strengths of the new methodology are demonstrated through analyses of three real cancer datasets and four simulation studies. The new cost-sensitive metric proved better performance in its ability to identify the best ordinal classifier for a given analysis. The performance metric devised in this study provides a comprehensive tool for comparative analysis of multiple (and competing) ordinal classifiers. Consideration of the tradeoff between accuracy and misclassification cost in decisions regarding ordinal classification problems is imperative in real-world application. The work presented here is a precursor to the possibility of incorporating the proposed metric into a prediction modeling algorithm for ordinal data as a means of integrating misclassification cost in final model selection.

Dietary Iodine Sufficiency and Moderate Insufficiency in the Lactating Mother and Nursing Infant: A Computational Perspective.

The Institute of Medicine recommends that lactating women ingest 290 µg iodide/d and a nursing infant, less than two years of age, 110 µg/d. The World Health Organization, United Nations Children's Fund, and International Council for the Control of Iodine Deficiency Disorders recommend population maternal and infant urinary iodide concentrations ≥ 100 µg/L to ensure iodide sufficiency. For breast milk, researchers have proposed an iodide concentration range of 150-180 µg/L indicates iodide sufficiency for the mother and infant, however no national or international guidelines exist for breast milk iodine concentration. For the first time, a lactating woman and nursing infant biologically based model, from delivery to 90 days postpartum, was constructed to predict maternal and infant urinary iodide concentration, breast milk iodide concentration, the amount of iodide transferred in breast milk to the nursing infant each day and maternal and infant serum thyroid hormone kinetics. The maternal and infant models each consisted of three sub-models, iodide, thyroxine (T4), and triiodothyronine (T3). Using our model to simulate a maternal intake of 290 µg iodide/d, the average daily amount of iodide ingested by the nursing infant, after 4 days of life, gradually increased from 50 to 101 µg/day over 90 days postpartum. The predicted average lactating mother and infant urinary iodide concentrations were both in excess of 100 µg/L and the predicted average breast milk iodide concentration, 157 µg/L. The predicted serum thyroid hormones (T4, free T4 (fT4), and T3) in both the nursing infant and lactating mother were indicative of euthyroidism. The model was calibrated using serum thyroid hormone concentrations for lactating women from the United States and was successful in predicting serum T4 and fT4 levels (within a factor of two) for lactating women in other countries. T3 levels were adequately predicted. Infant serum thyroid hormone levels were adequately predicted for most data. For moderate iodide deficient conditions, where dietary iodide intake may range from 50 to 150 µg/d for the lactating mother, the model satisfactorily described the iodide measurements, although with some variation, in urine and breast milk. Predictions of serum thyroid hormones in moderately iodide deficient lactating women (50 µg/d) and nursing infants did not closely agree with mean reported serum thyroid hormone levels, however, predictions were usually within a factor of two. Excellent agreement between prediction and observation was obtained for a recent moderate iodide deficiency study in lactating women. Measurements included iodide levels in urine of infant and mother, iodide in breast milk, and serum thyroid hormone levels in infant and mother. A maternal iodide intake of 50 µg/d resulted in a predicted 29-32% reduction in serum T4 and fT4 in nursing infants, however the reduced serum levels of T4 and fT4 were within most of the published reference intervals for infant. This biologically based model is an important first step at integrating the rapid changes that occur in the thyroid system of the nursing newborn in order to predict adverse outcomes from exposure to thyroid acting chemicals, drugs, radioactive materials or iodine deficiency.

Quantitative global sensitivity analysis of a biologically based dose-response pregnancy model for the thyroid endocrine system.

A deterministic biologically based dose-response model for the thyroidal system in a near-term pregnant woman and the fetus was recently developed to evaluate quantitatively thyroid hormone perturbations. The current work focuses on conducting a quantitative global sensitivity analysis on this complex model to identify and characterize the sources and contributions of uncertainties in the predicted model output. The workflow and methodologies suitable for computationally expensive models, such as the Morris screening method and Gaussian Emulation processes, were used for the implementation of the global sensitivity analysis. Sensitivity indices, such as main, total and interaction effects, were computed for a screened set of the total thyroidal system descriptive model input parameters. Furthermore, a narrower sub-set of the most influential parameters affecting the model output of maternal thyroid hormone levels were identified in addition to the characterization of their overall and pair-wise parameter interaction quotients. The characteristic trends of influence in model output for each of these individual model input parameters over their plausible ranges were elucidated using Gaussian Emulation processes. Through global sensitivity analysis we have gained a better understanding of the model behavior and performance beyond the domains of observation by the simultaneous variation in model inputs over their range of plausible uncertainties. The sensitivity analysis helped identify parameters that determine the driving mechanisms of the maternal and fetal iodide kinetics, thyroid function and their interactions, and contributed to an improved understanding of the system modeled. We have thus demonstrated the use and application of global sensitivity analysis for a biologically based dose-response model for sensitive life-stages such as pregnancy that provides richer information on the model and the thyroidal system modeled compared to local sensitivity analysis.

An Iterative Leave-One-Out Approach to Outlier Detection in RNA-Seq Data.

The discrete data structure and large sequencing depth of RNA sequencing (RNA-seq) experiments can often generate outlier read counts in one or more RNA samples within a homogeneous group. Thus, how to identify and manage outlier observations in RNA-seq data is an emerging topic of interest. One of the main objectives in these research efforts is to develop statistical methodology that effectively balances the impact of outlier observations and achieves maximal power for statistical testing. To reach that goal, strengthening the accuracy of outlier detection is an important precursor. Current outlier detection algorithms for RNA-seq data are executed within a testing framework and may be sensitive to sparse data and heavy-tailed distributions. Therefore, we propose a univariate algorithm that utilizes a probabilistic approach to measure the deviation between an observation and the distribution generating the remaining data and implement it within in an iterative leave-one-out design strategy. Analyses of real and simulated RNA-seq data show that the proposed methodology has higher outlier detection rates for both non-normalized and normalized negative binomial distributed data.