Francesco Marabita

A beta-mixture quantile normalization method for correcting probe design bias in Illumina Infinium 450 k DNA methylation data

Motivation: The Illumina Infinium 450 k DNA Methylation Beadchip is a prime candidate technology for Epigenome-Wide Association Studies (EWAS). However, a difficulty associated with these beadarrays is that probes come in two different designs, characterized by widely different DNA methylation distributions and dynamic range, which may bias downstream analyses. A key statistical issue is therefore how best to adjust for the two different probe designs. Results: Here we propose a novel model-based intra-array normalization strategy for 450 k data, called BMIQ (Beta MIxture Quantile dilation), to adjust the beta-values of type2 design probes into a statistical distribution characteristic of type1 probes. The strategy involves application of a three-state beta-mixture model to assign probes to methylation states, subsequent transformation of probabilities into quantiles and finally a methylation-dependent dilation transformation to preserve the monotonicity and continuity of the data. We validate our method on cell-line data, fresh frozen and paraffin-embedded tumour tissue samples and demonstrate that BMIQ compares favourably with two competing methods. Specifically, we show that BMIQ improves the robustness of the normalization procedure, reduces the technical variation and bias of type2 probe values and successfully eliminates the type1 enrichment bias caused by the lower dynamic range of type2 probes. BMIQ will be useful as a preprocessing step for any study using the Illumina Infinium 450 k platform. Availability: BMIQ is freely available from http://code.google.com/p/bmiq/. Contact: a.teschendorff@ucl.ac.uk Supplementary information: Supplementary data are available at Bioinformatics online

An evaluation of analysis pipelines for DNA methylation profiling using the Illumina HumanMethylation450 BeadChip platform

The proper identification of differentially methylated CpGs is central in most epigenetic studies. The Illumina HumanMethylation450 BeadChip is widely used to quantify DNA methylation; nevertheless, the design of an appropriate analysis pipeline faces severe challenges due to the convolution of biological and technical variability and the presence of a signal bias between Infinium I and II probe design types. Despite recent attempts to investigate how to analyze DNA methylation data with such an array design, it has not been possible to perform a comprehensive comparison between different bioinformatics pipelines due to the lack of appropriate data sets having both large sample size and sufficient number of technical replicates. Here we perform such a comparative analysis, targeting the problems of reducing the technical variability, eliminating the probe design bias and reducing the batch effect by exploiting two unpublished data sets, which included technical replicates and were profiled for DNA methylation either on peripheral blood, monocytes or muscle biopsies. We evaluated the performance of different analysis pipelines and demonstrated that: (1) it is critical to correct for the probe design type, since the amplitude of the measured methylation change depends on the underlying chemistry; (2) the effect of different normalization schemes is mixed, and the most effective method in our hands were quantile normalization and Beta Mixture Quantile dilation (BMIQ); (3) it is beneficial to correct for batch effects. In conclusion, our comparative analysis using a comprehensive data set suggests an efficient pipeline for proper identification of differentially methylated CpGs using the Illumina 450K arrays.

Genetic variation in the interleukin-28B gene is not associated with fibrosis progression in patients with chronic hepatitis C and known date of infection†

Polymorphisms in the interleukin‐28B (IL28B) region are associated with spontaneous and treatment‐induced viral clearance in hepatitis C virus (HCV) infection. Nevertheless, it is unknown whether genetic variation at the IL28B locus influences the natural history of chronic HCV infection. Thus, we asked whether an association between IL28B polymorphisms and liver fibrosis progression existed. We studied 247 consecutive patients with chronic HCV, an accurate estimate of the date of infection, and a liver biopsy performed before any treatment. No patient had a history of alcohol abuse or coinfection with other viruses. We assessed the role of rs8099917 and rs12979860 polymorphisms and the effect of host and environmental factors on fibrosis progression. Blood transfusion (75%) was the main modality of infection. Median age at infection was 21 years, and median interval between infection and liver biopsy was 25 years. One hundred twenty‐nine patients (52%) were infected by HCV‐1, 74 (30%) by HCV‐2, 34 (14%) by HCV‐3, and 10 (4%) by HCV‐4. Bridging fibrosis/cirrhosis (Ishak ≥4) was detected in 24% of patients. Age at infection had a marked effect on fibrosis progression by both a linear model and Cox proportional‐hazard regression (P < 2E‐16). A 12.1% increase in the hazard of advanced fibrosis was estimated for each additional year at infection, suggesting that this was the major explanatory variable in this cohort. Male gender (P < 0.05), HCV genotype 3 (P < 0.001) and steatosis (P < 0.05) were also associated with faster fibrosis progression. Conversely, the two IL28B polymorphisms had no impact on fibrosis progression. Conclusion: In HCV patients with a known date of infection, IL28B genotype was not associated with fibrosis progression rate or with the risk of developing advanced liver fibrosis. (HEPATOLOGY 2011;)

Normalization of circulating microRNA expression data obtained by quantitative real-time RT-PCR

The high-throughput analysis of microRNAs (miRNAs) circulating within the blood of healthy and diseased individuals is an active area of biomarker research. Whereas quantitative real-time reverse transcription polymerase chain reaction (qPCR)-based methods are widely used, it is yet unresolved how the data should be normalized. Here, we show that a combination of different algorithms results in the identification of candidate reference miRNAs that can be exploited as normalizers, in both discovery and validation phases. Using the methodology considered here, we identify normalizers that are able to reduce nonbiological variation in the data and we present several case studies, to illustrate the relevance in the context of physiological or pathological scenarios. In conclusion, the discovery of stable reference miRNAs from high-throughput studies allows appropriate normalization of focused qPCR assays.

An integrative analysis reveals coordinated reprogramming of the epigenome and the transcriptome in human skeletal muscle after training

Regular endurance exercise training induces beneficial functional and health effects in human skeletal muscle. The putative contribution to the training response of the epigenome as a mediator between genes and environment has not been clarified. Here we investigated the contribution of DNA methylation and associated transcriptomic changes in a well-controlled human intervention study. Training effects were mirrored by significant alterations in DNA methylation and gene expression in regions with a homogeneous muscle energetics and remodeling ontology. Moreover, a signature of DNA methylation and gene expression separated the samples based on training and gender. Differential DNA methylation was predominantly observed in enhancers, gene bodies and intergenic regions and less in CpG islands or promoters. We identified transcriptional regulator binding motifs of MRF, MEF2 and ETS proteins in the proximity of the changing sites. A transcriptional network analysis revealed modules harboring distinct ontologies and, interestingly, the overall direction of the changes of methylation within each module was inversely correlated to expression changes. In conclusion, we show that highly consistent and associated modifications in methylation and expression, concordant with observed health-enhancing phenotypic adaptations, are induced by a physiological stimulus.

Identification of serum and tissue micro-RNA expression profiles in different stages of inflammatory bowel disease

The altered expression of micro‐RNA (miRNA) has been associated with Crohns disease (CD) and ulcerative colitis (UC). The aim of this study was to establish specific miRNA expression patterns in the serum and mucosa of inflammatory bowel disease (IBD) patients (UC and CD with colonic involvement) at different stages of the disease. Serum and biopsies from nine active CD (aCD), nine inactive CD (iCD), nine active UC (aUC) and nine inactive UC (iUC) and serum from 33 healthy subjects were collected. Up to 700 miRNAs were evaluated by the TaqMan® human miRNA array. The ΔCt values were obtained using the mean expression values of all expressed miRNAs in a given sample as a normalization factor for miRNA real‐time quantitative polymerase chain reaction data. The levels of serum miRNAs in CD and UC patients were different to healthy subjects. Thirteen serum miRNAs were expressed commonly in CD and UC patients. Two miRNAs were higher and four miRNAs were lower in the serum of aCD than iCD. No serum miRNA was regulated exclusively in aUC compared with iUC patients. Four miRNAs were higher and three miRNAs were lower in the mucosa of aCD than iCD. Two miRNAs were higher and three miRNAs were lower in the mucosa of aUC than iUC. No serum miRNAs coincided with tissue miRNAs in aCD and aUC patients. Our results suggest the existence of specific miRNA expression patterns associated with IBD and their different stages and support the utility of miRNA as possible biomarkers. This pilot study needs to be validated in a large prospective cohort.

Intracellular Modulation, Extracellular Disposal and Serum Increase of MiR-150 Mark Lymphocyte Activation

Activated lymphocytes release nano-sized vesicles (exosomes) containing microRNAs that can be monitored in the bloodstream. We asked whether elicitation of immune responses is followed by release of lymphocyte-specific microRNAs. We found that, upon activation in vitro, human and mouse lymphocytes down-modulate intracellular miR-150 and accumulate it in exosomes. In vivo, miR-150 levels increased significantly in serum of humans immunized with flu vaccines and in mice immunized with ovalbumin, and this increase correlated with elevation of antibody titers. Immunization of immune-deficient mice, lacking MHCII, resulted neither in antibody production nor in elevation of circulating miR-150. This study provides proof of concept that serum microRNAs can be detected, with minimally invasive procedure, as biomarkers of vaccination and more in general of adaptive immune responses. Furthermore, the prompt reduction of intracellular level of miR-150, a key regulator of mRNAs critical for lymphocyte differentiation and functions, linked to its release in the external milieu suggests that the selective extracellular disposal of microRNAs can be a rapid way to regulate gene expression during lymphocyte activation.

Serum microRNAs as novel biomarkers for primary sclerosing cholangitis and cholangiocarcinoma.

The diagnosis of primary sclerosing cholangitis (PSC) is difficult due to the lack of sensitive and specific biomarkers, as is the early diagnosis of cholangiocarcinoma (CC), a complication of PSC. The aim of this study was to identify specific serum miRNAs as diagnostic biomarkers for PSC and CC. The levels of 667 miRNAs were evaluated in 90 human serum samples (30 PSC, 30 CC and 30 control subjects) to identify disease‐associated candidate miRNAs (discovery phase). The deregulated miRNAs were validated in an independent cohort of 140 samples [40 PSC, 40 CC, 20 primary biliary cirrhosis (PBC) and 40 controls]. Receiver operating characteristic (ROC) curves were established and only miRNAs with an area under the curve (AUC) > 0·70 were considered useful as biomarkers. In the discovery phase we identified the following: 21 miRNAs expressed differentially in PSC, 33 in CC and 26 in both in comparison to control subjects as well as 24 miRNAs expressed differentially between PSC and CC. After the validation phase, miR‐200c was found to be expressed differentially in PSC versus controls, whereas miR‐483‐5p and miR‐194 showed deregulated expression in CC compared with controls. We also demonstrate a difference in the expression of miR‐222 and miR‐483‐5p in CC versus PSC. Combination of these specific miRNAs further improved the specificity and accuracy of diagnosis. This study provides a basis for the use of miRNAs as biomarkers for the diagnosis of PSC and CC.

Systems Medicine: from molecular features and models to the clinic in COPD

Background and hypothesisChronic Obstructive Pulmonary Disease (COPD) patients are characterized by heterogeneous clinical manifestations and patterns of disease progression. Two major factors that can be used to identify COPD subtypes are muscle dysfunction/wasting and co-morbidity patterns. We hypothesized that COPD heterogeneity is in part the result of complex interactions between several genes and pathways. We explored the possibility of using a Systems Medicine approach to identify such pathways, as well as to generate predictive computational models that may be used in clinic practice.Objective and methodOur overarching goal is to generate clinically applicable predictive models that characterize COPD heterogeneity through a Systems Medicine approach. To this end we have developed a general framework, consisting of three steps/objectives: (1) feature identification, (2) model generation and statistical validation, and (3) application and validation of the predictive models in the clinical scenario. We used muscle dysfunction and co-morbidity as test cases for this framework.ResultsIn the study of muscle wasting we identified relevant features (genes) by a network analysis and generated predictive models that integrate mechanistic and probabilistic models. This allowed us to characterize muscle wasting as a general de-regulation of pathway interactions. In the co-morbidity analysis we identified relevant features (genes/pathways) by the integration of gene-disease and disease-disease associations. We further present a detailed characterization of co-morbidities in COPD patients that was implemented into a predictive model. In both use cases we were able to achieve predictive modeling but we also identified several key challenges, the most pressing being the validation and implementation into actual clinical practice.ConclusionsThe results confirm the potential of the Systems Medicine approach to study complex diseases and generate clinically relevant predictive models. Our study also highlights important obstacles and bottlenecks for such approaches (e.g. data availability and normalization of frameworks among others) and suggests specific proposals to overcome them.

Transcriptomics and methylomics of CD4-positive T cells in arsenic-exposed women

Arsenic, a carcinogen with immunotoxic effects, is a common contaminant of drinking water and certain food worldwide. We hypothesized that chronic arsenic exposure alters gene expression, potentially by altering DNA methylation of genes encoding central components of the immune system. We therefore analyzed the transcriptomes (by RNA sequencing) and methylomes (by target-enrichment next-generation sequencing) of primary CD4-positive T cells from matched groups of four women each in the Argentinean Andes, with fivefold differences in urinary arsenic concentrations (median concentrations of urinary arsenic in the lower- and high-arsenic groups: 65 and 276 μg/l, respectively). Arsenic exposure was associated with genome-wide alterations of gene expression; principal component analysis indicated that the exposure explained 53% of the variance in gene expression among the top variable genes and 19% of 28,351 genes were differentially expressed (false discovery rate <0.05) between the exposure groups. Key genes regulating the immune system, such as tumor necrosis factor alpha and interferon gamma, as well as genes related to the NF-kappa-beta complex, were significantly downregulated in the high-arsenic group. Arsenic exposure was associated with genome-wide DNA methylation; the high-arsenic group had 3% points higher genome-wide full methylation (>80% methylation) than the lower-arsenic group. Differentially methylated regions that were hyper-methylated in the high-arsenic group showed enrichment for immune-related gene ontologies that constitute the basic functions of CD4-positive T cells, such as isotype switching and lymphocyte activation and differentiation. In conclusion, chronic arsenic exposure from drinking water was related to changes in the transcriptome and methylome of CD4-positive T cells, both genome wide and in specific genes, supporting the hypothesis that arsenic causes immunotoxicity by interfering with gene expression and regulation.