Using RNA sample titrations to assess microarray platform performance and normalization techniques

Ameliorative effect of the oil components derived from sweet potato shochu on impaired short-term memory in mice after amyloid β25-35 injection

Sweet potato shochu oil is a by-product of shochu production and usually discarded although some physiological functions are considered. In this study, we investigated the effects of shochu oil on short-term memory using a murine model of spontaneous alternating behavior induced by the intracerebroventricular (ICV) administration of amyloid β_25-35 (Aβ_25-35). Mice were orally administered shochu oil for 15 days. Experiments with a Y-maze model revealed that the Aβ_25-35 caused a significant decrease in spontaneous alternation behavior, and supplementation with shochu oil significantly improved this behavior. DNA microarray analysis revealed that the administration of shochu oil downregulated the expression of S100a9 and Ptgs2, which reportedly exacerbate amyloid β deposition in Alzheimer's disease. The administration of shochu oil upregulated the expression of Dnaja1 and PP2A, which is typically downregulated in Alzheimer's disease. These data suggest that shochu oil possible ameliorates on impaired short-term memory in mice after amyloid β_25-35 injection, as indicated by its effects on improving spontaneous alternation behavior and modulating the expressions of related genes.

Low-Level Erbium-Doped Yttrium Aluminum Garnet Laser Irradiation Induced Alteration of Gene Expression in Osteogenic Cells from Rat Calvariae

Objective: The aim of this study was to investigate the effect of low-level erbium-doped yttrium aluminum garnet (Er:YAG) laser irradiation on gene expression in osteogenic cells from rat calvariae. Background: Previous studies showed beneficial effects of laser irradiation on bone-related cells. However, few studies have examined the gene expression alteration by laser irradiation on osteogenic cells in a calcified condition. Materials and methods: Osteogenic cells were prepared by culturing rat calvarial osteoblast-like cells in osteoinductive medium for 21 days. The cells at the bottom of the culture dish were irradiated with Er:YAG laser (wavelength: 2.94 μm, energy density: 3.1 and 8.2 J/cm²) positioned at distance of 25 cm. Lactate dehydrogenase (LDH) assay of the irradiated cells was performed. After screening for genes related to bone formation, mechanotransduction, and thermal effect by quantitative polymerase chain reaction (qPCR), gene expression at 3 h after 3.1 J/cm² irradiation was comprehensively analyzed using microarray. Results: No dramatical increase in surface temperature and LDH activities after laser irradiation were observed. Sost expression was significantly reduced at 3 h after 3.1 J/cm² irradiation. Bcar1 and Hspa1a expression was significantly increased following 8.2 J/cm² irradiation. Microarray analysis identified 116 differentially expressed genes. Gene set enrichment analysis showed enrichment of histone H3-K9 methylation and modification gene sets. Conclusions: Er:YAG laser irradiation, especially at 3.1 J/cm², showed positive effect on the expression of genes related to bone formation in osteogenic cells, without inducing significant cell damage. These findings may represent critical mechanisms of early bone formation after Er:YAG laser irradiation.

A blind and independent benchmark study for detecting differeally methylated regions in plants

MOTIVATION

Bisulfite sequencing (BS-seq) is a state-of-the-art technique for investigating methylation of the DNA to gain insights into the epigenetic regulation. Several algorithms have been published for identification of differentially methylated regions (DMRs). However, the performances of the individual methods remain unclear and it is difficult to optimally select an algorithm in application settings.

RESULTS

We analyzed BS-seq data from four plants covering three taxonomic groups. We first characterized the data using multiple summary statistics describing methylation levels, coverage, noise, as well as frequencies, magnitudes and lengths of methylated regions. Then, simulated datasets with most similar characteristics to real experimental data were created. Seven different algorithms (metilene, methylKit, MOABS, DMRcate, Defiant, BSmooth, MethylSig) for DMR identification were applied and their performances were assessed. A blind and independent study design was chosen to reduce bias and to derive practical method selection guidelines. Overall, metilene had superior performance in most settings. Data attributes, such as coverage and spread of the DMR lengths, were found to be useful for selecting the best method for DMR detection. A decision tree to select the optimal approach based on these data attributes is provided. The presented procedure might serve as a general strategy for deriving algorithm selection rules tailored to demands in specific application settings.

AVAILABILITY AND IMPLEMENTATION

Scripts that were used for the analyses and that can be used for prediction of the optimal algorithm are provided at https://github.com/kreutz-lab/DMR-DecisionTree. Simulated and experimental data are available at https://doi.org/10.6084/m9.figshare.11619045.

CONTACT

ckreutz@imbi.uni-freiburg.de.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Flåm S, Leithaug M, Dalland M, Rayner S, Lie BA, Gilfillan GD. An extension to: Systematic assessment of commercially available low-input miRNA library preparation kits

High-throughput sequencing has emerged as the favoured method to study microRNA (miRNA) expression, but biases introduced during library preparation have been reported. We recently compared the performance (sensitivity, reliability, titration response and differential expression) of six commercially-available kits on synthetic miRNAs and human RNA, where library preparation was performed by the vendors. We hereby supplement this study with data from two further commonly used kits (NEBNext, NEXTflex) whose manufacturers initially declined to participate. NEXTflex demonstrated the highest sensitivity, which may reflect its use of partially-randomized adapter sequences, but overall performance was lower than the QIAseq and TailorMix kits. NEBNext showed intermediate performance. We reaffirm that biases are kit specific, complicating the comparison of miRNA datasets generated using different kits.

Microarray Normalization Revisited for Reproducible Breast Cancer Biomarkers

Precision medicine for breast cancer relies on biomarkers to select therapies. However, the reliability of biomarkers drawn from gene expression arrays has been questioned and calls for reassessment, in particular for large datasets. We revisit widely used data-normalization procedures and evaluate differences in outcome in order to pinpoint the most reliable reprocessing methods biomarkers can be based upon. We generated a database of 3753 breast cancer patients out of 38 studies by downloading and curating patient samples from NCBI-GEO. As gene-expression biomarkers, we select the assessment of receptor status and breast cancer subtype classification. Each normalization procedure is applied separately, and biomarkers are then evaluated for each patient. Differences between normalization pipelines are quantified as percentages of patients having outcomes different for each pipeline. Some normalization procedures lead to quite consistent biomarkers, differing only in 1-2% of patients. Other normalization procedures—some of them have been used in many clinical studies—end up with distrusting discrepancies (10% and more). A good deal of doubt regarding the reliability of microarrays may root in the haphazard application of inadequate preprocessing pipelines. Several modes of batch corrections are evaluated regarding a possible improvement of receptor prediction from gene expression versus the golden standard of immunohistochemistry. Finally, we nominate those normalization methods yielding consistent and trustable results. Adequate bioinformatics data preprocessing is key and crucial for any subsequent statistics to arrive at trustable results. We conclude with a suggestion for future bioinformatics development to further increase the reliability of cancer biomarkers.

Effects of Low-Level Er:YAG Laser Irradiation on Proliferation and Calcification of Primary Osteoblast-Like Cells Isolated From Rat Calvaria

Several reports have shown that the photo-bio-modulation of cells by various lasers has favorable biological effects. However, the effects of low-level Er:YAG laser irradiation on osteoblasts remain unclear. The purpose of this study was to evaluate the effects of low-level Er:YAG laser irradiation on proliferation and osteogenic differentiation of primary osteoblast-like cells isolated from the calvariae of 3-5-day-old Wistar rats. Cells were irradiated by Er:YAG laser at energy fluences of 2.2, 3.3, and 4.3 J/cm2, respectively. After irradiation, cell surface temperatures were measured and cell proliferation was evaluated by flow cytometry and CCK-8. Calcification was evaluated by measuring areas of Alizarin red S staining after 7, 14, and 21 days culture in osteoinductive medium. Gene expression in non-irradiated and laser-irradiated cells was evaluated by qPCR at 3, 6, and 12 h, as well as 1, 3, 7, and 14 days after irradiation. Microarray analysis was performed to comprehensively evaluate the gene expression of non-irradiated and irradiated cells at 3.3 J/cm2 at 6 h after irradiation. No pronounced increase of cell surface temperature was induced by irradiation. Irradiation did not affect osteoblast-like cell proliferation. Osteoblast-like cell calcification was significantly increased 7 days after Er:YAG laser irradiation at 3.3 J/cm2. Bglap expression was significantly increased in cells irradiated at 3.3 J/cm2 6 h post-irradiation. Microarray analysis showed that irradiation at 3.3 J/cm2 caused an upregulation of inflammation-related genes and downregulation of Wisp2. Gene set enrichment analysis also clarified enrichment of inflammation-related and Notch signaling gene sets. In conclusion, low-level Er:YAG laser irradiation at 3.3 J/cm2 enhanced calcification of primary osteoblast-like cells via enhanced Bglap expression and enriched Notch signaling.

Combined effect of canagliflozin and exercise training on high-fat diet-fed mice

Sodium-glucose cotransporter 2 inhibitors (SGLT2is) have been reported to improve obesity, diabetes, and nonalcoholic fatty liver disease (NAFLD) in addition to exercise training, whereas the combined effects remain to be elucidated fully. We investigated the effect of the combination of the SGLT2i canagliflozin (CAN) and exercise training in high-fat diet-induced obese mice. High-fat diet-fed mice were housed in normal cages (sedentary; Sed) or wheel cages (WCR) with or without CAN (0.03% of diet) for 4 wk. The effects on obesity, glucose metabolism, and hepatic steatosis were evaluated in four groups (Control/Sed, Control/WCR, CAN/Sed, and CAN/WCR). Numerically additive improvements were found in body weight, body fat mass, blood glucose, glucose intolerance, insulin resistance, and the fatty liver of the CAN/WCR group, whereas CAN increased food intake and reduced running distance. Exercise training alone, CAN alone, or both did not change the weight of skeletal muscle, but microarray analysis showed that each resulted in a characteristic change of gene expression in gastrocnemius muscle. In particular, in the CAN/WCR group, there was acceleration of the angiogenesis pathway and suppression of the adipogenesis pathway compared with the CAN/Sed group. In conclusion, the combination of an SGLT2i and exercise training improves obesity, insulin resistance, and NAFLD in an additive manner. Changes of gene expression in skeletal muscle may contribute, at least in part, to the improvement of obesity and insulin sensitivity.

Systematic Identification of Housekeeping Genes Possibly Used as References in Caenorhabditis elegans by Large-Scale Data Integration

For accurate gene expression quantification, normalization of gene expression data against reliable reference genes is required. It is known that the expression levels of commonly used reference genes vary considerably under different experimental conditions, and therefore, their use for data normalization is limited. In this study, an unbiased identification of reference genes in Caenorhabditis elegans was performed based on 145 microarray datasets (2296 gene array samples) covering different developmental stages, different tissues, drug treatments, lifestyle, and various stresses. As a result, thirteen housekeeping genes (rps-23, rps-26, rps-27, rps-16, rps-2, rps-4, rps-17, rpl-24.1, rpl-27, rpl-33, rpl-36, rpl-35, and rpl-15) with enhanced stability were comprehensively identified by using six popular normalization algorithms and RankAggreg method. Functional enrichment analysis revealed that these genes were significantly overrepresented in GO terms or KEGG pathways related to ribosomes. Validation analysis using recently published datasets revealed that the expressions of newly identified candidate reference genes were more stable than the commonly used reference genes. Based on the results, we recommended using rpl-33 and rps-26 as the optimal reference genes for microarray and rps-2 and rps-4 for RNA-sequencing data validation. More importantly, the most stable rps-23 should be a promising reference gene for both data types. This study, for the first time, successfully displays a large-scale microarray data driven genome-wide identification of stable reference genes for normalizing gene expression data and provides a potential guideline on the selection of universal internal reference genes in C. elegans, for quantitative gene expression analysis.

Using the circulating proteome to assess type I interferon activity in systemic lupus erythematosus

Type I interferon (IFN) drives pathology in systemic lupus erythematosus (SLE) and can be tracked via IFN-inducible transcripts in blood. Here, we examined whether measurement of circulating proteins, which enter the bloodstream from inflamed tissues, also offers insight into global IFN activity. Using a novel protocol we generated 1,132 aptamer-based protein measurements from anti-dsDNA^pos SLE blood samples and derived an IFN protein signature (IFNPS) that approximates the IFN 21-gene signature (IFNGS). Of 82 patients with SLE, IFNPS was elevated for 89% of IFNGS-high patients (49/55) and 26% of IFNGS-low patients (7/27). IFNGS-high/IFNPS-high patients exhibited activated NK, CD4, and CD8 T cells, while IFNPS-high only patients did not. IFNPS correlated with global disease activity in lymphopenic and non-lymphopenic patients and decreased following type I IFN neutralisation with anifrolumab in the SLE phase IIb study, MUSE. In summary, we developed a protein signature that reflects IFNGS and identifies a new subset of patients with SLE who have IFN activity.

Systematic assessment of commercially available low-input miRNA library preparation kits

High-throughput sequencing is increasingly favoured to assay the presence and abundance of microRNAs (miRNAs) in biological samples, even from low RNA amounts, and a number of commercial vendors now offer kits that allow miRNA sequencing from sub-nanogram (ng) inputs. Although biases introduced during library preparation have been documented, the relative performance of current reagent kits has not been investigated in detail. Here, six commercial kits capable of handling <100ng total RNA input were used for library preparation, performed by kit manufactures, on synthetic miRNAs of known quantities and human total RNA samples. We compared the performance of miRNA detection sensitivity, reliability, titration response and the ability to detect differentially expressed miRNAs. In addition, we assessed the use of unique molecular identifiers (UMI) sequence tags in one kit. We observed differences in detection sensitivity and ability to identify differentially expressed miRNAs between the kits, but none were able to detect the full repertoire of synthetic miRNAs. The reliability within the replicates of all kits was good, while larger differences were observed between the kits, although none could accurately quantify the relative levels of the majority of miRNAs. UMI tags, at least within the input ranges tested, offered little advantage to improve data utility. In conclusion, biases in miRNA abundance are heavily influenced by the kit used for library preparation, suggesting that comparisons of datasets prepared by different procedures should be made with caution. This article is intended to assist researchers select the most appropriate kit for their experimental conditions.

Practical determination of LODP (limit of detection for microarray platform) for the evaluation of microarray platforms

The performance indicator called limit of detection for microarray platform (LODP) was defined in ISO 16578:2013. The methods to determine practical LODP were explored. In general, + 3 SD of the background is used as the signal strength of limit of detection and criteria for dividing positive and negative results. Since the negative signal had been defined differently for each microarray platform, signals obtained from Non-Probe Spots (NPS) installed on the microarrays were defined as the "background" of microarrays. LODP was determined as the lowest concentration of which the average signal exceeded Avg. + 3 SD of the background (NPS) and the signal was significantly different from those of the lower and higher adjacent concentration points measured with a diluted series of reference materials. For reliable qualitative analysis, the positive results can be defined as signals higher than those corresponding to LODP and negative results as lower signals, without determining limit of detection for all target probes. The use of LODP also enables comparisons of platform performances without checking sequence dependencies, and assists to select reliable and fitting platforms for experimental purposes.

Evaluation of bone healing following Er:YAG laser ablation in rat calvaria compared with bur drilling

The Er:YAG laser is currently used for bone ablation. However, the effect of Er:YAG laser irradiation on bone healing remains unclear. The aim of this study was to investigate bone healing following ablation by laser irradiation as compared with bur drilling. Rat calvarial bone was ablated using Er:YAG laser or bur with water coolant. Er:YAG laser effectively ablated bone without major thermal changes. In vivo micro-computed tomography analysis revealed that laser irradiation showed significantly higher bone repair ratios than bur drilling. Scanning electron microscope analysis showed more fibrin deposition on laser-ablated bone surfaces. Microarray analysis followed by gene set enrichment analysis revealed that IL6/JAK/STAT3 signaling and inflammatory response gene sets were enriched in bur-drilled bone at 6 hours, whereas the E2F targets gene set was enriched in laser-irradiated bone. Additionally, Hspa1a and Dmp1 expressions were increased and Sost expression was decreased in laser-irradiated bone compared with bur-drilled bone. In granulation tissue formed after laser ablation, Alpl and Gblap expressions increased compared to bur-drilled site. Immunohistochemistry showed that osteocalcin-positive area was increased in the laser-ablated site. These results suggest that Er:YAG laser might accelerate early new bone formation with advantageous surface changes and cellular responses for wound healing, compared with bur-drilling.

Endotoxemia by Porphyromonas gingivalis Injection Aggravates Non-alcoholic Fatty Liver Disease, Disrupts Glucose/Lipid Metabolism, and Alters Gut Microbiota in Mice

Many risk factors related to the development of non-alcoholic fatty liver disease (NAFLD) have been proposed, including the most well-known of diabetes and obesity as well as periodontitis. As periodontal pathogenic bacteria produce endotoxins, periodontal treatment can result in endotoxemia. The aim of this study was to investigate the effects of intravenous, sonicated Porphyromonas gingivalis (Pg) injection on glucose/lipid metabolism, liver steatosis, and gut microbiota in mice. Endotoxemia was induced in C57BL/6J mice (8 weeks old) by intravenous injection of sonicated Pg; Pg was deactivated but its endotoxin remained. The mice were fed a high-fat diet and administered sonicated Pg (HFPg) or saline (HFco) injections for 12 weeks. Liver steatosis, glucose metabolism, and gene expression in the liver were evaluated. 16S rRNA gene sequencing with metagenome prediction was performed on the gut microbiota. Compared to HFco mice, HFPg mice exhibited impaired glucose tolerance and insulin resistance along with increased liver steatosis. Liver microarray analysis demonstrated that 1278 genes were differentially expressed between HFco and HFPg mice. Gene set enrichment analysis showed that fatty acid metabolism, hypoxia, and TNFα signaling via NFκB gene sets were enriched in HFPg mice. Although sonicated Pg did not directly reach the gut, it changed the gut microbiota and decreased bacterial diversity in HFPg mice. Metagenome prediction in the gut microbiota showed enriched citrate cycle and carbon fixation pathways in prokaryotes. Overall, intravenous injection of sonicated Pg caused impaired glucose tolerance, insulin resistance, and liver steatosis in mice fed high-fat diets. Thus, blood infusion of Pg contributes to NAFLD and alters the gut microbiota.

Distinct molecular pathways in ovarian endometrioid adenocarcinoma with concurrent endometriosis

Women with endometriosis, a benign growth of endometrial tissue outside the uterine cavity, are at increased risk of specific histotypes of epithelial ovarian cancer, such as ovarian endometrioid adenocarcinoma (OEA). Women with OEA who have endometriosis at time of surgical staging demonstrate improved clinical prognosis compared to women with OEA without evidence of endometriosis. However, the molecular contributions of the endometriotic tumor microenvironment to these ovarian cancers remain poorly understood. As a starting point, we used a platform for genome-wide transcriptomic profiling to compare specimens of OEA from women with and without concurrent endometriosis and benign reproductive tract tissues, including proliferative endometrium and typical and atypical endometrioma samples (n = 20). Principle component analysis revealed distinct clustering between benign and malignant samples as well as malignant samples with and without concurrent endometriosis. Examination of gene signatures revealed that OEA with concurrent endometriosis contained a unique molecular signature compared to OEA without concurrent endometriosis, distinguished by 682 unique genes differentially expressed (fold change < or >1.5, p < 0.01). Bioinformatic analysis of these differentially expressed gene products using ingenuity pathway analysis revealed activation of NFkB signaling, an inflammatory signaling pathway constitutively active in endometriosis. DAVID functional annotation clustering further revealed enrichment in RAS signaling as both cytoskeleton organization and GTPase regulator activity relied heavily on RAS protein signal transduction. Gene set enrichment analysis highlighted immune and inflammatory nodes involved in OEA with concurrent endometriosis. These observations provide novel resources for understanding molecular subtleties potentially involved in OEA within the context of the endometriotic tumor microenvironment.

Apparent bias toward long gene misregulation in MeCP2 syndromes disappears after controlling for baseline variations

Recent studies have suggested that genes longer than 100 kb are more likely to be misregulated in neurological diseases associated with synaptic dysfunction, such as autism and Rett syndrome. These length-dependent transcriptional changes are modest in MeCP2-mutant samples, but, given the low sensitivity of high-throughput transcriptome profiling technology, here we re-evaluate the statistical significance of these results. We find that the apparent length-dependent trends previously observed in MeCP2 microarray and RNA-sequencing datasets disappear after estimating baseline variability from randomized control samples. This is particularly true for genes with low fold changes. We find no bias with NanoString technology, so this long gene bias seems to be particular to polymerase chain reaction amplification-based platforms. In contrast, authentic long gene effects, such as those caused by topoisomerase inhibition, can be detected even after adjustment for baseline variability. We conclude that accurate characterization of length-dependent (or other) trends requires establishing a baseline from randomized control samples.

Effect of Porphyromonas gingivalis infection in the placenta and umbilical cord in pregnant mice with low birth weight

OBJECTIVE

Growing evidence indicates an association between periodontitis and delivery outcome; however, the mechanism is unclear. This study aimed to investigate the influence of Porphyromonas gingivalis (Pg) infection on delivery outcome in mice.

MATERIALS AND METHODS

Bacteremia was induced in pregnant Slc:ICR mice (8 weeks old) by intravenous injection of Pg. Mice were randomly divided into a control group (CO), and those receiving Pg injection at gestational day 1 (GD1), gestational day 15 (GD15) or every day (ED). Delivery outcome, Pg infection, and gene expression in the placenta and umbilical cord were evaluated.

RESULTS

Birth weight was lower in the ED and GD15 groups than in the CO group. A remarkable increase in anti-Pg IgG antibody was observed in the ED and GD1 groups, although Pg was not detected in the placenta or umbilical cord. mRNA expression of Tnfα and Il6 in the placenta, and Hif1α in the umbilical cord, was significantly increased in the ED group. Microarray analysis of the umbilical cord revealed increased expression of several genes including Orm1, Mgl2, Rps6ka3 and Trim15 in the ED group.

CONCLUSIONS

Pg infection during the third trimester caused low birth weight and inflammation in the placenta and umbilical cord.

Specific alteration of gene expression profile in rats by treatment with thyroid toxicants that inhibit thyroid hormone synthesis

Transcriptomics technologies have been used for risk assessment of chemicals, mainly to predict the modes of action (MOAs) of chemicals or identify biomarkers. Transcriptomics data may also be helpful to understand MOAs of chemicals at the molecular level in more detail. As an example of the known MOAs, there are two MOAs of thyroid toxicity: inhibition of thyroid hormone synthesis ("direct" effect) and hypermetabolism of thyroid hormone by enzyme induction in liver ("indirect" effect). In the present study, global profiles of gene expression were analyzed in rats treated with chemicals acting directly on the thyroid (thyroid peroxidase inhibitors such as propylthiouracil and methimazole) and chemicals acting indirectly on the thyroid (hepatic enzyme inducers such as phenobarbital and pregnenolone-16α-carbonitrile) using microarrays. Using a subtraction method between these two types of chemicals, we identified characteristic gene expression changes on the thyroid hormone synthesis pathway by direct-acting chemicals. Based on the functions of these genes, alterations of their expression seem to indicate the results of thyroid peroxidase inhibition, and might be helpful in more accurate evaluation of MOAs for thyroid toxicity.

A high-throughput 3' UTR reporter screening identifies microRNA interactomes of cancer genes

Introduction

Despite the established contribution of deregulated microRNA (miRNA) function to carcinogenesis, relatively few miRNA-cancer gene interactions have been validated, making it difficult to appreciate the true complexity of miRNA-cancer gene regulatory networks.

Results

In this effort, we identify miRNA interactomes of 17 well-established cancer genes, involved in various cancer types, through a miRNome-wide 3’ UTR reporter screening. Using a novel and performant strategy for high-throughput screening data analysis, we identify 390 interactions, quadrupling the size of the known miRNA interactome for the cancer genes under investigation. Clear enrichments of established and predicted interactions underscore the validity of the interactome data set. Interactomes appear to be primarily driven by canonical binding site interactions. Nonetheless, non-canonical binding sites, such as offset 6mer and seed-mismatched or G:U wobble sites, also have regulatory activity, albeit clearly less pronounced. Furthermore, we observe enhanced regulation in the presence of 3’ supplementary pairing for both canonical and non-canonical binding sites.

Conclusions

Altogether, the cancer gene-miRNA interactome data set represents a unique resource that will aid in the unraveling of regulatory miRNA networks and the dynamic regulation of key protein-coding cancer genes. In addition, it uncovers aspects of the functional miRNA binding site’s architecture and the relative contributions of different binding site types.

Summarizing performance for genome scale measurement of miRNA: reference samples and metrics

BACKGROUND

The potential utility of microRNA as biomarkers for early detection of cancer and other diseases is being investigated with genome-scale profiling of differentially expressed microRNA. Processes for measurement assurance are critical components of genome-scale measurements. Here, we evaluated the utility of a set of total RNA samples, designed with between-sample differences in the relative abundance of miRNAs, as process controls.

RESULTS

Three pure total human RNA samples (brain, liver, and placenta) and two different mixtures of these components were evaluated as measurement assurance control samples on multiple measurement systems at multiple sites and over multiple rounds. In silico modeling of mixtures provided benchmark values for comparison with physical mixtures. Biomarker development laboratories using next-generation sequencing (NGS) or genome-scale hybridization assays participated in the study and returned data from the samples using their routine workflows. Multiplexed and single assay reverse-transcription PCR (RT-PCR) was used to confirm in silico predicted sample differences. Data visualizations and summary metrics for genome-scale miRNA profiling assessment were developed using this dataset, and a range of performance was observed. These metrics have been incorporated into an online data analysis pipeline and provide a convenient dashboard view of results from experiments following the described design. The website also serves as a repository for the accumulation of performance values providing new participants in the project an opportunity to learn what may be achievable with similar measurement processes.

CONCLUSIONS

The set of reference samples used in this study provides benchmark values suitable for assessing genome-scale miRNA profiling processes. Incorporation of these metrics into an online resource allows laboratories to periodically evaluate their performance and assess any changes introduced into their measurement process.

Candidate genes responsible for early key events of phenobarbital-promoted mouse hepatocellular tumorigenesis based on differentiation of regulating genes between wild type mice and humanized chimeric mice

Phenobarbital (PB) is a nongenotoxic hepatocellular carcinogen in rodents. PB induces hepatocellular tumors by activating the constitutive androstane receptor (CAR). Some previous research has suggested the possible involvement of epigenetic regulation in PB-promoted hepatocellular tumorigenesis, but the details of its molecular mechanism are not fully understood. In the present study, comprehensive analyses of DNA methylation, hydroxymethylation and gene expression using microarrays were performed in mouse hepatocellular adenomas induced by a single 90 mg kg-1 intraperitoneal injection dose of diethylnitrosamine (DEN) followed by 500 ppm PB in the diet for 27 weeks. DNA modification and expression of hundreds of genes are coordinately altered in PB-induced mouse hepatocellular adenomas. Of these, gene network analysis showed alterations of CAR signaling and tumor development-related genes. Pathway enrichment analysis revealed that differentially methylated or hydroxymethylated genes belong mainly to pathways involved in development, immune response and cancer cells in contrast to differentially expressed genes belonging primarily to the cell cycle. Furthermore, overlap was evaluated between the genes with altered expression levels with 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC) alterations in mouse hepatocellular adenoma induced by DEN/PB and the genes with altered expression levels in the liver of CD-1 mice or humanized chimeric mice treated with PB for 7 days. With the integration of transcriptomic and epigenetic approaches, we detected candidate genes responsible for early key events of PB-promoted mouse hepatocellular tumorigenesis. Interestingly, these genes did not overlap with genes altered by the PB treatment of humanized chimeric mice, thus suggesting a species difference between the effects of PB in mouse and human hepatocytes.

Periodontal pathogenic bacteria, Aggregatibacter actinomycetemcomitans affect non-alcoholic fatty liver disease by altering gut microbiota and glucose metabolism

Increasing evidence indicates that periodontitis affects non-alcoholic fatty liver disease (NAFLD). We examined the relationship between periodontal bacterial infection and clinical/biochemical parameters in 52 NAFLD patients. Anti-Aggregatibacter actinomycetemcomitans (Aa) antibody titers correlated positively with visceral fat, fasting plasma insulin, and HOMA-IR; and negatively with the liver/spleen ratio. C57BL/6J mice (8-weeks-old) were given Aa or saline (control) for 6 weeks, and were fed either normal chow (NCAa, NCco) or high-fat diet (HFAa and HFco). NCAa and HFAa mice presented impaired glucose tolerance and insulin resistance compared to control mice. HFAa mice showed higher hepatic steatosis than HFco animals. Liver microarray analysis revealed that 266 genes were differentially expressed between NCAa and NCco mice. Upregulated genes in Aa-administrated mice were enriched for glucagon signaling pathway, adipocytokine signaling pathway and insulin resistance. Consistently, plasma glucagon concentration was higher in NCAa mice. In addition, Akt phosphorylation was lower in the liver of NCAa/HFAa than in NCco/HFco mice. Based on 16S rRNA sequencing, Aa administration changed composition of the gut microbiota. Metagenome prediction in gut microbiota showed upregulation of fatty acid biosynthesis and downregulation of fatty acid degradation in Aa-administered mice. Thus, infection with Aa affects NAFLD by altering the gut microbiota and glucose metabolism.

Early gestational maternal low-protein diet diminishes hepatic response to fasting in young adult male mice

Maternal low-protein (MLP) diet can lead to hepatic steatosis, which only develops with ageing. It is still unclear whether the young offspring show any signs of past exposure to prenatal adverse conditions. We hypothesized that early nutritional insult would first affect the dynamic responsiveness to nutritional challenges rather than the static state. We analyzed the transcriptome and metabolome profiles of the hepatic response to fasting/refeeding in young male mice offspring to identify changes induced by early gestational MLP diet. Restricted MLP exposure strictly to early gestation was achieved by the embryo transfer method. As a result, the fasting-induced upregulation of genes related to long-chain fatty acid metabolism and of stress response genes related to protein folding were significantly diminished in MLP pups. Lipid profiling after fasting showed that the hepatic signature of triacylglycerols was shifted to longer acyl-chains and higher saturation by the MLP diet. Bioinformatic analyses suggested that these phenomenological changes may be partially linked to the peroxisome proliferator activated receptor α (PPARα) pathway. Taken together, early gestational MLP diet affected the hepatic dynamic response to nutritional stress in seemingly healthy young offspring, accompanied with partial deterioration of PPARα action.

Evaluation of the chondroprotective action of N-acetylglucosamine in a rat experimental osteoarthritis model

It has been demonstrated that oral administration of N-acetylglucosamine (GlcNAc) alleviates the symptoms of osteoarthritis (OA). The aim of the present study was to elucidate the molecular mechanisms for the chondroprotective action of GlcNAc in OA. Biomarkers for type II collagen degradation and synthesis were evaluated, as were histopathological changes, using a rat anterior cruciate ligament transection (ACLT)-induced OA model. Changes in the expression of genes in the cartilage were assessed via DNA microarray and reverse transcription-quantitative polymerase chain reaction (RT-qPCR). The results indicated that ACLT induced histopathological changes of articular cartilage, whereas oral administration of GlcNAc (1,000 mg/kg/day for 28 days) significantly suppressed these changes. Additionally, GlcNAc significantly decreased levels of a type II collagen degradation marker in sera compared with that in the ACLT group, although there were no significant changes in the levels of a type II collagen synthesis marker. Furthermore, DNA microarray and reverse transcription-quantitative polymerase chain reaction results demonstrated that GlcNAc treatment downregulated the expression of periostin, which is likely involved in the degradation of cartilage, whereas GlcNAc upregulated the expression of lipocalin 2, which is involved in the regulation of chondrocyte proliferation and differentiation. In conclusion, the results of the present study suggest that GlcNAc is able to suppress the histopathological changes induced by OA and exhibits a chondroprotective action by inhibiting type II collagen degradation in the articular cartilage, possibly via modulation of the expression of inflammatory and chondroprotective molecules, including periostin and lipocalin 2.

Variation-preserving normalization unveils blind spots in gene expression profiling

RNA-Seq and gene expression microarrays provide comprehensive profiles of gene activity, but lack of reproducibility has hindered their application. A key challenge in the data analysis is the normalization of gene expression levels, which is currently performed following the implicit assumption that most genes are not differentially expressed. Here, we present a mathematical approach to normalization that makes no assumption of this sort. We have found that variation in gene expression is much larger than currently believed, and that it can be measured with available assays. Our results also explain, at least partially, the reproducibility problems encountered in transcriptomics studies. We expect that this improvement in detection will help efforts to realize the full potential of gene expression profiling, especially in analyses of cellular processes involving complex modulations of gene expression.

In silico Pathway Activation Network Decomposition Analysis (iPANDA) as a method for biomarker development

Signalling pathway activation analysis is a powerful approach for extracting biologically relevant features from large-scale transcriptomic and proteomic data. However, modern pathway-based methods often fail to provide stable pathway signatures of a specific phenotype or reliable disease biomarkers. In the present study, we introduce the in silico Pathway Activation Network Decomposition Analysis (iPANDA) as a scalable robust method for biomarker identification using gene expression data. The iPANDA method combines precalculated gene coexpression data with gene importance factors based on the degree of differential gene expression and pathway topology decomposition for obtaining pathway activation scores. Using Microarray Analysis Quality Control (MAQC) data sets and pretreatment data on Taxol-based neoadjuvant breast cancer therapy from multiple sources, we demonstrate that iPANDA provides significant noise reduction in transcriptomic data and identifies highly robust sets of biologically relevant pathway signatures. We successfully apply iPANDA for stratifying breast cancer patients according to their sensitivity to neoadjuvant therapy.

Syndecan-4 as a biomarker to predict clinical outcome for glioblastoma multiforme treated with WT1 peptide vaccine

AIM

In cancer immunotherapy, biomarkers are important for identification of responsive patients. This study was aimed to find biomarkers that predict clinical outcome of WT1 peptide vaccination.

MATERIALS & METHODS

Candidate genes that were expressed differentially between long- and short-term survivors were identified by cDNA microarray analysis of peripheral blood mononuclear cells that were extracted from 30 glioblastoma patients (discovery set) prior to vaccination and validated by quantitative RT-PCR using discovery set and different 23 patients (validation set).

RESULTS

SDC-4 mRNA expression levels distinguished between the long- and short-term survivors: 1-year survival rates were 64.0 and 18.5% in SDC4-low and -high patients, respectively.

CONCLUSION

SDC-4 is a novel predictive biomarker for the efficacy of WT1 peptide vaccine.

The FDA's Experience with Emerging Genomics Technologies-Past, Present, and Future

The rapid advancement of emerging genomics technologies and their application for assessing safety and efficacy of FDA-regulated products require a high standard of reliability and robustness supporting regulatory decision-making in the FDA. To facilitate the regulatory application, the FDA implemented a novel data submission program, Voluntary Genomics Data Submission (VGDS), and also to engage the stakeholders. As part of the endeavor, for the past 10 years, the FDA has led an international consortium of regulatory agencies, academia, pharmaceutical companies, and genomics platform providers, which was named MicroArray Quality Control Consortium (MAQC), to address issues such as reproducibility, precision, specificity/sensitivity, and data interpretation. Three projects have been completed so far assessing these genomics technologies: gene expression microarrays, whole genome genotyping arrays, and whole transcriptome sequencing (i.e., RNA-seq). The resultant studies provide the basic parameters for fit-for-purpose application of these new data streams in regulatory environments, and the solutions have been made available to the public through peer-reviewed publications. The latest MAQC project is also called the SEquencing Quality Control (SEQC) project focused on next-generation sequencing. Using reference samples with built-in controls, SEQC studies have demonstrated that relative gene expression can be measured accurately and reliably across laboratories and RNA-seq platforms. Besides prediction performance comparable to microarrays in clinical settings and safety assessments, RNA-seq is shown to have better sensitivity for low expression and reveal novel transcriptomic features. Future effort of MAQC will be focused on quality control of whole genome sequencing and targeted sequencing.

Integration of flow studies for robust selection of mechanoresponsive genes

Blood flow is an essential contributor to plaque growth, composition and initiation. It is sensed by endothelial cells, which react to blood flow by expressing > 1000 genes. The sheer number of genes implies that one needs genomic techniques to unravel their response in disease. Individual genomic studies have been performed but lack sufficient power to identify subtle changes in gene expression. In this study, we investigated whether a systematic meta-analysis of available microarray studies can improve their consistency. We identified 17 studies using microarrays, of which six were performed in vivo and 11 in vitro. The in vivo studies were disregarded due to the lack of the shear profile. Of the in vitro studies, a cross-platform integration of human studies (HUVECs in flow cells) showed high concordance (> 90 %). The human data set identified > 1600 genes to be shear responsive, more than any other study and in this gene set all known mechanosensitive genes and pathways were present. A detailed network analysis indicated a power distribution (e. g. the presence of hubs), without a hierarchical organisation. The average cluster coefficient was high and further analysis indicated an aggregation of 3 and 4 element motifs, indicating a high prevalence of feedback and feed forward loops, similar to prokaryotic cells. In conclusion, this initial study presented a novel method to integrate human-based mechanosensitive studies to increase its power. The robust network was large, contained all known mechanosensitive pathways and its structure revealed hubs, and a large aggregate of feedback and feed forward loops.

Comprehensive evaluation of AmpliSeq transcriptome, a novel targeted whole transcriptome RNA sequencing methodology for global gene expression analysis

Background

Whole transcriptome sequencing (RNA-seq) represents a powerful approach for whole transcriptome gene expression analysis. However, RNA-seq carries a few limitations, e.g., the requirement of a significant amount of input RNA and complications led by non-specific mapping of short reads. The Ion AmpliSeq™ Transcriptome Human Gene Expression Kit (AmpliSeq) was recently introduced by Life Technologies as a whole-transcriptome, targeted gene quantification kit to overcome these limitations of RNA-seq. To assess the performance of this new methodology, we performed a comprehensive comparison of AmpliSeq with RNA-seq using two well-established next-generation sequencing platforms (Illumina HiSeq and Ion Torrent Proton). We analyzed standard reference RNA samples and RNA samples obtained from human induced pluripotent stem cell derived cardiomyocytes (hiPSC-CMs).

Results

Using published data from two standard RNA reference samples, we observed a strong concordance of log2 fold change for all genes when comparing AmpliSeq to Illumina HiSeq (Pearson’s r = 0.92) and Ion Torrent Proton (Pearson’s r = 0.92). We used ROC, Matthew’s correlation coefficient and RMSD to determine the overall performance characteristics. All three statistical methods demonstrate AmpliSeq as a highly accurate method for differential gene expression analysis. Additionally, for genes with high abundance, AmpliSeq outperforms the two RNA-seq methods. When analyzing four closely related hiPSC-CM lines, we show that both AmpliSeq and RNA-seq capture similar global gene expression patterns consistent with known sources of variations.

Conclusions

Our study indicates that AmpliSeq excels in the limiting areas of RNA-seq for gene expression quantification analysis. Thus, AmpliSeq stands as a very sensitive and cost-effective approach for very large scale gene expression analysis and mRNA marker screening with high accuracy.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-2270-1) contains supplementary material, which is available to authorized users.

Standardization efforts enabling next-generation sequencing and microarray based biomarkers for precision medicine

Microarrays and next-generation sequencing technologies have been increasingly employed in biomedical research. However, before they can be reliably used as clinical biomarker tests, standardization and quality control measures need to be developed to ensure their analytical validity. This review summarizes community-wide efforts such as the MicroArray and Sequencing Quality Control (MAQC/SEQC) project which have identified factors influencing the performance of these technologies. Consequently, consensus-based standards and well-documented best practices have been developed to improve the quality of scientific research, and reference materials and reference datasets have been made available for evaluating the technical proficiency in future studies. These efforts have built the foundation on which the translational application of genomics based technologies can help realize precision medicine.

Using mixtures of biological samples as process controls for RNA-sequencing experiments

Background

Genome-scale “-omics” measurements are challenging to benchmark due to the enormous variety of unique biological molecules involved. Mixtures of previously-characterized samples can be used to benchmark repeatability and reproducibility using component proportions as truth for the measurement. We describe and evaluate experiments characterizing the performance of RNA-sequencing (RNA-Seq) measurements, and discuss cases where mixtures can serve as effective process controls.

Results

We apply a linear model to total RNA mixture samples in RNA-seq experiments. This model provides a context for performance benchmarking. The parameters of the model fit to experimental results can be evaluated to assess bias and variability of the measurement of a mixture. A linear model describes the behavior of mixture expression measures and provides a context for performance benchmarking. Residuals from fitting the model to experimental data can be used as a metric for evaluating the effect that an individual step in an experimental process has on the linear response function and precision of the underlying measurement while identifying signals affected by interference from other sources. Effective benchmarking requires well-defined mixtures, which for RNA-Seq requires knowledge of the post-enrichment ‘target RNA’ content of the individual total RNA components. We demonstrate and evaluate an experimental method suitable for use in genome-scale process control and lay out a method utilizing spike-in controls to determine enriched RNA content of total RNA in samples.

Conclusions

Genome-scale process controls can be derived from mixtures. These controls relate prior knowledge of individual components to a complex mixture, allowing assessment of measurement performance. The target RNA fraction accounts for differential selection of RNA out of variable total RNA samples. Spike-in controls can be utilized to measure this relationship between target RNA content and input total RNA. Our mixture analysis method also enables estimation of the proportions of an unknown mixture, even when component-specific markers are not previously known, whenever pure components are measured alongside the mixture.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1912-7) contains supplementary material, which is available to authorized users.

High-throughput RNA profiling via up-front sample parallelization

We describe a method called modular, early-tagged amplification (META) RNA profiling that can quantify a broad panel of microRNAs or mRNAs simultaneously across many samples and requires far less sequence depth than existing digital profiling technologies. The method assigns quantitative tags during reverse transcription to permit up-front sample pooling before competitive amplification and deep sequencing. This simple, scalable and inexpensive approach improves the practicality of large-scale gene expression studies.

A comprehensive assessment of RNA-seq accuracy, reproducibility and information content by the Sequencing Quality Control Consortium

We present primary results from the Sequencing Quality Control (SEQC) project, coordinated by the US Food and Drug Administration. Examining Illumina HiSeq, Life Technologies SOLiD and Roche 454 platforms at multiple laboratory sites using reference RNA samples with built-in controls, we assess RNA sequencing (RNA-seq) performance for junction discovery and differential expression profiling and compare it to microarray and quantitative PCR (qPCR) data using complementary metrics. At all sequencing depths, we discover unannotated exon-exon junctions, with >80% validated by qPCR. We find that measurements of relative expression are accurate and reproducible across sites and platforms if specific filters are used. In contrast, RNA-seq and microarrays do not provide accurate absolute measurements, and gene-specific biases are observed for all examined platforms, including qPCR. Measurement performance depends on the platform and data analysis pipeline, and variation is large for transcript-level profiling. The complete SEQC data sets, comprising >100 billion reads (10Tb), provide unique resources for evaluating RNA-seq analyses for clinical and regulatory settings.

Distinguishing between maternally-mediated and directly embryotoxic modes-of-action for postimplantation loss induced in rats by 2-amino-2-methylpropanol

In rats, 2-amino-2-methylpropanol (AMP) caused an increase in postimplantation loss in an oral reproductive/developmental toxicity screening assay but not in a dermal developmental toxicity assay. Studies were performed to characterize the mode of action and determine whether the postimplantation loss was a result of direct embryotoxicity or a maternally mediated effect. The studies identified that the postimplantation loss occurs shortly after implantation, has a steep dose response with a clear threshold, requires exposure to AMP for a period of approximately 2-3 weeks prior to gestation and does not involve direct embryo toxicity. The uterine histopathology and gene array analysis of decidual swellings suggested AMP acts via a maternally mediated mechanism affecting the ability of the uterus to support an implanted embryo. Since the postimplantation loss occurs only at maternally toxic doses, the implications for human risk assessment are discussed.

Comparison of gene expression microarray data with count-based RNA measurements informs microarray interpretation

Background

Although numerous investigations have compared gene expression microarray platforms, preprocessing methods and batch correction algorithms using constructed spike-in or dilution datasets, there remains a paucity of studies examining the properties of microarray data using diverse biological samples. Most microarray experiments seek to identify subtle differences between samples with variable background noise, a scenario poorly represented by constructed datasets. Thus, microarray users lack important information regarding the complexities introduced in real-world experimental settings. The recent development of a multiplexed, digital technology for nucleic acid measurement enables counting of individual RNA molecules without amplification and, for the first time, permits such a study.

Results

Using a set of human leukocyte subset RNA samples, we compared previously acquired microarray expression values with RNA molecule counts determined by the nCounter Analysis System (NanoString Technologies) in selected genes. We found that gene measurements across samples correlated well between the two platforms, particularly for high-variance genes, while genes deemed unexpressed by the nCounter generally had both low expression and low variance on the microarray. Confirming previous findings from spike-in and dilution datasets, this “gold-standard” comparison demonstrated signal compression that varied dramatically by expression level and, to a lesser extent, by dataset. Most importantly, examination of three different cell types revealed that noise levels differed across tissues.

Conclusions

Microarray measurements generally correlate with relative RNA molecule counts within optimal ranges but suffer from expression-dependent accuracy bias and precision that varies across datasets. We urge microarray users to consider expression-level effects in signal interpretation and to evaluate noise properties in each dataset independently.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-649) contains supplementary material, which is available to authorized users.

Evaluation of quantitative miRNA expression platforms in the microRNA quality control (miRQC) study

MicroRNAs are important negative regulators of protein-coding gene expression and have been studied intensively over the past years. Several measurement platforms have been developed to determine relative miRNA abundance in biological samples using different technologies such as small RNA sequencing, reverse transcription-quantitative PCR (RT-qPCR) and (microarray) hybridization. In this study, we systematically compared 12 commercially available platforms for analysis of microRNA expression. We measured an identical set of 20 standardized positive and negative control samples, including human universal reference RNA, human brain RNA and titrations thereof, human serum samples and synthetic spikes from microRNA family members with varying homology. We developed robust quality metrics to objectively assess platform performance in terms of reproducibility, sensitivity, accuracy, specificity and concordance of differential expression. The results indicate that each method has its strengths and weaknesses, which help to guide informed selection of a quantitative microRNA gene expression platform for particular study goals.

Association of bone morphogenetic protein 6 with exocrine gland dysfunction in patients with Sjögren's syndrome and in mice

OBJECTIVE

Primary Sjögren's syndrome (SS) is characterized by autoimmune activation and loss of function in secretory epithelia. The present study was undertaken to investigate and characterize changes in the epithelia associated with the loss of gland function in primary SS.

METHODS

To identify changes in epithelial gene expression, custom microarrays were probed with complementary RNA (cRNA) isolated from minor salivary glands (MSGs) of female patients with primary SS who had low focus scores and low salivary flow rates, and the results were compared with those obtained using cRNA from the MSGs of sex-matched healthy volunteers. The effect of bone morphogenetic protein 6 (BMP-6) on salivary gland function was tested using adeno-associated virus-mediated gene transfer to the salivary glands of C57BL/6 mice.

RESULTS

A significant increase in expression of BMP-6 was observed in RNA isolated from SS patients compared with healthy volunteers. Overexpression of BMP-6 locally in the salivary or lacrimal glands of mice resulted in the loss of fluid secretion as well as changes in the connective tissue of the salivary gland. Assessment of the fluid movement in either isolated acinar cells from mice overexpressing BMP-6 or a human salivary gland cell line cultured with BMP-6 revealed a loss in volume regulation in these cells. Lymphocytic infiltration in the submandibular gland of BMP-6 vector-treated mice was increased. No significant changes in the production of proinflammatory cytokines or autoantibodies associated with SS (anti-Ro/SSA and anti-La/SSB) were found after BMP-6 overexpression.

CONCLUSION

In addition to identifying BMP-6 expression in association with xerostomia and xerophthalmia in primary SS, the present results suggest that BMP-6-induced salivary and lacrimal gland dysfunction in primary SS is independent of the autoantibodies and immune activation associated with the disease.

Evaluation of Different Normalization and Analysis Procedures for Illumina Gene Expression Microarray Data Involving Small Changes

While Illumina microarrays can be used successfully for detecting small gene expression changes due to their high degree of technical replicability, there is little information on how different normalization and differential expression analysis strategies affect outcomes. To evaluate this, we assessed concordance across gene lists generated by applying different combinations of normalization strategy and analytical approach to two Illumina datasets with modest expression changes. In addition to using traditional statistical approaches, we also tested an approach based on combinatorial optimization. We found that the choice of both normalization strategy and analytical approach considerably affected outcomes, in some cases leading to substantial differences in gene lists and subsequent pathway analysis results. Our findings suggest that important biological phenomena may be overlooked when there is a routine practice of using only one approach to investigate all microarray datasets. Analytical artefacts of this kind are likely to be especially relevant for datasets involving small fold changes, where inherent technical variation-if not adequately minimized by effective normalization-may overshadow true biological variation. This report provides some basic guidelines for optimizing outcomes when working with Illumina datasets involving small expression changes.

Molecular biology in breast cancer: should molecular classifiers be assessed by conventional tools or by gene expression arrays?

Breast cancer is a complex disease, with heterogeneous presentations and clinical courses. Standard clinico-pathological parameters, relying on single gene or protein characterization determined with sometimes poorly-reproducible technologies, have shown limitations in the classification of the disease and in the prediction of individual patient outcomes and responses to therapy. Gene-expression profiling has revealed great potential to accurately classify breast cancer and define patient prognosis and prediction to anti-cancer therapy. Nevertheless, the performance of molecular classifiers remains sub-optimal, and both technical and conceptual improvements are needed. It is likely that determining the ideal strategy for tailoring treatment of breast cancer will require a more systematic, structured and multi-dimensional approach than in the past. Besides implementing cutting-edge technologies to detect genetic and epigenetic cancer alterations, the future of breast cancer research will in all probability rely on the innovative and multilevel integration of molecular profiles with clinical parameters of the disease and patient-related factors.

NanoStringNorm: an extensible R package for the pre-processing of NanoString mRNA and miRNA data

Motivation: The NanoString nCounter Platform is a new and promising technology for measuring nucleic acid abundances. It has several advantages over PCR-based techniques, including avoidance of amplification, direct sequence interrogation and digital detection for absolute quantification. These features minimize aspects of experimental error and hold promise for dealing with challenging experimental conditions such as archival formalin-fixed paraffin-embedded samples. However, systematic inter-sample technical artifacts caused by variability in sample preservation, bio-molecular extraction and platform fluctuations must be removed to ensure robust data.

Results: To facilitate this process and to address these issues for NanoString datasets, we have written a pre-processing package called NanoStringNorm in the R statistical language. Key features include an extensible environment for method comparison and new algorithm development, integrated gene and sample diagnostics, and facilitated downstream statistical analysis. The package is open-source, is available through the CRAN package repository, includes unit-tests to ensure numerical accuracy, and provides visual and numeric diagnostics.

Availability:http://cran.r-project.org/web/packages/NanoStringNorm

Contact:paul.boutros@oicr.on.ca

Supplementary information:Supplementary data are available at Bioinformatics online.

Impact of the spotted microarray preprocessing method on fold-change compression and variance stability

Background

The standard approach for preprocessing spotted microarray data is to subtract the local background intensity from the spot foreground intensity, to perform a log2 transformation and to normalize the data with a global median or a lowess normalization. Although well motivated, standard approaches for background correction and for transformation have been widely criticized because they produce high variance at low intensities. Whereas various alternatives to the standard background correction methods and to log2 transformation were proposed, impacts of both successive preprocessing steps were not compared in an objective way.

Results

In this study, we assessed the impact of eight preprocessing methods combining four background correction methods and two transformations (the log2 and the glog), by using data from the MAQC study. The current results indicate that most preprocessing methods produce fold-change compression at low intensities. Fold-change compression was minimized using the Standard and the Edwards background correction methods coupled with a log2 transformation. The drawback of both methods is a high variance at low intensities which consequently produced poor estimations of the p-values. On the other hand, effective stabilization of the variance as well as better estimations of the p-values were observed after the glog transformation.

Conclusion

As both fold-change magnitudes and p-values are important in the context of microarray class comparison studies, we therefore recommend to combine the Edwards correction with a hybrid transformation method that uses the log2 transformation to estimate fold-change magnitudes and the glog transformation to estimate p-values.

Evaluating methods for ranking differentially expressed genes applied to microArray quality control data

Background

Statistical methods for ranking differentially expressed genes (DEGs) from gene expression data should be evaluated with regard to high sensitivity, specificity, and reproducibility. In our previous studies, we evaluated eight gene ranking methods applied to only Affymetrix GeneChip data. A more general evaluation that also includes other microarray platforms, such as the Agilent or Illumina systems, is desirable for determining which methods are suitable for each platform and which method has better inter-platform reproducibility.

Results

We compared the eight gene ranking methods using the MicroArray Quality Control (MAQC) datasets produced by five manufacturers: Affymetrix, Applied Biosystems, Agilent, GE Healthcare, and Illumina. The area under the curve (AUC) was used as a measure for both sensitivity and specificity. Although the highest AUC values can vary with the definition of "true" DEGs, the best methods were, in most cases, either the weighted average difference (WAD), rank products (RP), or intensity-based moderated t statistic (ibmT). The percentages of overlapping genes (POGs) across different test sites were mainly evaluated as a measure for both intra- and inter-platform reproducibility. The POG values for WAD were the highest overall, irrespective of the choice of microarray platform. The high intra- and inter-platform reproducibility of WAD was also observed at a higher biological function level.

Conclusion

These results for the five microarray platforms were consistent with our previous ones based on 36 real experimental datasets measured using the Affymetrix platform. Thus, recommendations made using the MAQC benchmark data might be universally applicable.

An adaptable method using human mixed tissue ratiometric controls for benchmarking performance on gene expression microarrays in clinical laboratories

BACKGROUND

Molecular biomarkers that are based on mRNA transcripts are being developed for the diagnosis and treatment of a number of diseases. DNA microarrays are one of the primary technologies being used to develop classifiers from gene expression data for clinically relevant outcomes. Microarray assays are highly multiplexed measures of comparative gene expression but have a limited dynamic range of measurement and show compression in fold change detection. To increase the clinical utility of microarrays, assay controls are needed that benchmark performance using metrics that are relevant to the analysis of genomic data generated with biological samples.

RESULTS

Ratiometric controls were prepared from commercial sources of high quality RNA from human tissues with distinctly different expression profiles and mixed in defined ratios. The samples were processed using six different target labeling protocols and replicate datasets were generated on high density gene expression microarrays. The area under the curve from receiver operating characteristic plots was calculated to measure diagnostic performance. The reliable region of the dynamic range was derived from log(2) ratio deviation plots made for each dataset. Small but statistically significant differences in diagnostic performance were observed between standardized assays available from the array manufacturer and alternative methods for target generation. Assay performance using the reliable range of comparative measurement as a metric was improved by adjusting sample hybridization conditions for one commercial kit.

CONCLUSIONS

Process improvement in microarray assay performance was demonstrated using samples prepared from commercially available materials and two metrics - diagnostic performance and the reliable range of measurement. These methods have advantages over approaches that use a limited set of external controls or correlations to reference sets, because they provide benchmark values that can be used by clinical laboratories to help optimize protocol conditions and laboratory proficiency with microarray assays.