Normalization is an essential step with considerable impact on high-throughput RNA sequencing (RNA-seq) data analysis. Although there are numerous methods for read count normalization, it remains a challenge to choose an optimal method due to multiple factors contributing to read count variability that affects the overall sensitivity and specificity. In order to properly determine the most appropriate normalization methods, it is critical to compare the performance and shortcomings of a representative set of normalization routines based on different dataset characteristics. Therefore, we set out to evaluate the performance of the commonly used methods (DESeq, TMM-edgeR, FPKM-CuffDiff, TC, Med UQ and FQ) and two new methods we propose: Med-pgQ2 and UQ-pgQ2 (per-gene normalization after per-sample median or upper-quartile global scaling). Our per-gene normalization approach allows for comparisons between conditions based on similar count levels. Using the benchmark Microarray Quality Control Project (MAQC) and simulated datasets, we performed differential gene expression analysis to evaluate these methods. When evaluating MAQC2 with two replicates, we observed that Med-pgQ2 and UQ-pgQ2 achieved a slightly higher area under the Receiver Operating Characteristic Curve (AUC), a specificity rate > 85%, the detection power > 92% and an actual false discovery rate (FDR) under 0.06 given the nominal FDR (≤0.05). Although the top commonly used methods (DESeq and TMM-edgeR) yield a higher power (>93%) for MAQC2 data, they trade off with a reduced specificity (<70%) and a slightly higher actual FDR than our proposed methods. In addition, the results from an analysis based on the qualitative characteristics of sample distribution for MAQC2 and human breast cancer datasets show that only our gene-wise normalization methods corrected data skewed towards lower read counts. However, when we evaluated MAQC3 with less variation in five replicates, all methods performed similarly. Thus, our proposed Med-pgQ2 and UQ-pgQ2 methods perform slightly better for differential gene analysis of RNA-seq data skewed towards lowly expressed read counts with high variation by improving specificity while maintaining a good detection power with a control of the nominal FDR level.

AIMTo investigate genetic factors that might help define which Crohn’s disease (CD) patients are likely to benefit from anti-tumor necrosis factor (TNF) therapy.METHODSThis was a prospective cohort study. Patients were recruited from a university digestive disease practice database. We included CD patients who received anti-TNF therapy, had available medical records (with information on treatment duration and efficacy) and who consented to participation. Patients with allergic reactions were excluded. Patients were grouped as ever-responders or non-responders. Genomic DNA was extracted from peripheral blood, and 7 single nucleotide polymorphisms (SNPs) were assessed. The main outcome measure (following exposure to the drug) was response to therapy. The patient genotypes were assessed as the predictors of outcome. Possible confounders and effect modifiers included age, gender, race, and socioeconomic status disease, as well as disease characteristics (such as Montreal criteria).RESULTS121 patients were included. Twenty-one were non-responders, and 100 were ever-responders. Fas ligand SNP (rs763110) genotype frequencies, TNF gene -308 SNP (rs1800629) genotype frequencies, and their combination, were significantly different between groups on multivariable analysis controlling for Montreal disease behavior and perianal disease. The odds of a patient with a Fas ligand CC genotype being a non-responder were four-fold higher as compared to a TC or TT genotype (P = 0.009, OR = 4.30, 95%CI: 1.45-12.80). The presence of the A (minor) TNF gene -308 allele correlated with three-fold higher odds of being a non-responder (P = 0.049, OR = 2.88, 95%CI: 1.01-8.22). Patients with the combination of the Fas ligand CC genotype and the TNF -308 A allele had nearly five-fold higher odds of being a non-responder (P = 0.015, OR = 4.76, 95%CI: 1.35-16.77). No difference was seen for the remaining SNPs.CONCLUSIONThe Fas-ligand SNP and TNF gene -308 SNP are associated with anti-TNF treatment response in CD and may help select patients likely to benefit from therapy.

Chronic arsenic exposure causes skin cancer, although the underlying molecular mechanisms are not well defined. Altered microRNA and mRNA expression likely play a pivotal role in carcinogenesis. Changes in genome-wide differential expression of miRNA and mRNA at 3 strategic time points upon chronic sodium arsenite (As3+) exposure were investigated in a well-validated HaCaT cell line model of arsenic-induced cutaneous squamous cell carcinoma (cSCC). Quadruplicate independent HaCaT cell cultures were exposed to 0 or 100 nM As3+ for up to 28-weeks (wk). Cell growth was monitored throughout the course of exposure and epithelial-mesenchymal transition (EMT) was examined employing immunoblot. Differentially expressed miRNA and mRNA profiles were generated at 7, 19, and 28-wk by RNA-seq, followed by identification of differentially expressed mRNA targets of differentially expressed miRNAs through expression pairing at each time point. Pathway analyses were performed for total differentially expressed mRNAs and for the miRNA targeted mRNAs at each time point. RNA-seq predictions were validated by immunoblot of selected target proteins. While the As3+-exposed cells grew slower initially, growth was equal to that of unexposed cells by 19-wk (transformation initiation), and exposed cells subsequently grew faster than passage-matched unexposed cells. As3+-exposed cells had undergone EMT at 28-wk. Pathway analyses demonstrate dysregulation of carcinogenesis-related pathways and networks in a complex coordinated manner at each time point. Immunoblot data largely corroborate RNA-seq predictions in the endoplasmic reticulum stress (ER stress) pathway. This study provides a detailed molecular picture of changes occurring during the arsenic-induced transformation of human keratinocytes.

Platelet count has been shown to be lower and mean platelet volume (MPV) to be higher in acute myocardial infarction (MI). However, it is not known whether these changes persist post-MI or if these measures are able to distinguish between acute thrombotic and non-thrombotic MI. Platelet count and MPV were measured in 80 subjects with acute MI (thrombotic and non-thrombotic) and stable coronary artery disease (CAD) at cardiac catheterization (acute phase) and at >3-month follow-up (quiescent phase). Subjects were stratified using stringent clinical, biochemical, histological, and angiographic criteria. Outcome measures were compared between groups by analysis of variance. Forty-seven subjects met criteria for acute MI with clearly defined thrombotic (n = 22) and non-thrombotic (n = 12) subsets. Fourteen subjects met criteria for stable CAD. No significant difference was observed in platelet count between subjects with acute MI and stable CAD at the acute or quiescent phase. MPV was higher in acute MI (9.18 ± 1.21) compared to stable CAD (8.13 ± 0.66; P = 0.003) at the acute phase but not at the quiescent phase (8.48 ± 0.58 vs 8.94 ± 1.42; P = 0.19). No difference in platelet count or MPV was detected between thrombotic and non-thrombotic subsets at acute or quiescent phases. The power to detect differences in these measures between thrombotic and non-thrombotic subsets was 58%. Higher MPV at the time of acute MI is not observed by 3 months post-MI (quiescent phase). Platelet count and MPV do not differ in subjects with thrombotic versus non-thrombotic MI. Further investigation is warranted to evaluate the utility of these measures in the diagnosis of acute MI.