Array Datasets Research Articles

Impact of Interleukin-6 –174 G>C Gene Promoter Polymorphism on Neuroblastoma

BackgroundCommon variants in DNA may predispose to onset and progression of neuroblastoma (NB). The genotype GG of single nucleotide polymorphism (SNP) rs1800795 (−174 G>C) in interleukin (IL)-6 promoter has been associated with lower survival of high-risk NB.ResultTo evaluate the impact of IL-6 SNP rs1800795 on disease risk and phenotype, we analyzed 326 Italian NB patients and 511 controls. Moreover, we performed in silico and quantitative Real Time (qRT)-PCR analyses to evaluate the influence of the SNP on gene expression in 198 lymphoblastoid cell lines (LCLs) and in 31 NB tumors, respectively. Kaplan-Meier analysis was used to verify the association between IL-6 gene expression and patient survival. We found that IL-6 SNP is not involved in susceptibility to NB development. However, our results show that a low frequency of genotype CC is significantly associated with a low overall survival, advanced stage, and high-risk phenotype. The in silico (p = 2.61×10−5) and qRT-PCR (p = 0.03) analyses showed similar trend indicating that the CC genotype is correlated with increased level of IL-6 expression. In report gene assay, we showed that the −174 C variant had a significantly increased transcriptional activity compared with G allele (p = 0.0006). Moreover, Kaplan-Meier analysis demonstrated that high levels of IL-6 are associated with poor outcome in children with NB in two independent gene expression array datasets.ConclusionsThe biological effect of SNP IL-6–174 G>C in relation to promotion of cancer progression is consistent with the observed decreased survival time. The present study suggests that SNP IL-6–174 G>C may be a useful marker for NB prognosis.

A Comparative Analysis of Rough Set Based Intelligent Techniques for Unsupervised Gene Selection

As the micro array databases increases in dimension and results in complexity, identifying the most informative genes is a challenging task. Such difficulty is often related to the huge number of genes with very few samples. Research in medical data mining addresses this problem by applying techniques from data mining and machine learning to the micro array datasets. In this paper Unsupervised Tolerance Rough Set based Quick Reduct (U-TRS-QR), a diverse feature selection algorithm, which extends the existing equivalent rough sets for unsupervised learning, is proposed. Genes selected by the proposed method leads to a considerably improved class predictions in wide experiments on two gene expression datasets: Brain Tumor and Colon Cancer. The results indicate consistent improvement among 12 classifiers.

Abstract PR11: Ovarian cancer cells activate the mesothelium in the peritoneal microenvironment promoting fibronectin production and the early steps of metastasis

Abstract The main site of ovarian cancer (OvCa) metastasis is the omentum, which is covered by a single layer of mesothelial cells overlaying an extracellular matrix interspersed with fibroblasts (mesothelium). A 3D model of the human peritoneal surface (in vitro), fragments of human omentum (ex vivo), and a mouse xenograft model (in vivo) of OvCa were used to investigate the early effects of cancer cells on the host stromal microenvironment. Our results reveal that fibronectin is overexpressed in the stroma of more than 93% of OvCa omental metastases. OvCa cells induce fibronectin matrix assembly in omental stromal cells, and stimulate fibronectin expression in mesothelial cells. The fibronectin produced by human mesothelial stromal cells promotes OvCa cell adhesion, invasion and proliferation. Inhibition of fibronectin production in mesothelial cells or omental surface cells using a fibronectin specific siRNA decreases OvCa cell adhesion, invasion and proliferation. Genetic knock-down of fibronectin in vitro and in vivo, using fibronectinfl/fl mice, also impaired OvCa cell adhesion/invasion and metastasis. Re-analysis of an Australian OvCa study gene array data set suggested that OvCa, which highly express fibronectin, show co-activation of TGFβ1/Rac1 dependent signaling pathways. Indeed, OvCa cells increase phosphorylation of Smad2/3 in mesothelial cells upon co-culture. Inhibition of TGFβ1 in OvCa cells or inhibition of Smad 3, as well as TGFβRI or TGFβRII, in mesothelial cells resulted in decreased fibronectin production. Moreover, Rac 1 activity was increased upon the co-culture of OvCa and mesothelial cells. Subsequent inhibition of Rac 1 in mesothelial cells led to a decrease in fibronectin secretion. Our data suggest that OvCa cells regulate fibronectin expression in mesothelial cells through a TGFβ1/Rac1 dependent pathway. In vitro and in vivo therapy with a fibronectin inhibitor, ATN161, a peptide derived from the synergy region of fibronectin, reduced OvCa invasion, proliferation and metastasis. In summary, we show that abdominally metastasizing cancer cells induce fibronectin production in the peritoneal and omental microenvironment and that inhibition of this fibronectin-response prevents the first steps of peritoneal metastasis. By understanding the mechanism of OvCa metastasis we will be better able to target cells in the tumor microenvironment (i.e. mesothelial cells, fibroblasts and macrophages) during early ovarian cancer metastasis. This abstract is also presented as Poster A60. Citation Format: Hilary A. Kenny, Chun-Yi Chiang, Andras Ladanyi, Joshy George, David D. Bowtell, Katja Gwin, Andrew Mazar, Ernst Lengyel. Ovarian cancer cells activate the mesothelium in the peritoneal microenvironment promoting fibronectin production and the early steps of metastasis. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: From Concept to Clinic; Sep 18-21, 2013; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2013;19(19 Suppl):Abstract nr PR11.

Tissue-specific isoform switch and DNA hypomethylation of the pyruvate kinase PKM gene in human cancers

The M2 isoform of pyruvate kinase (PKM2) plays an important role in aerobic glycolysis and is a mediator of the Warburg effect in tumors. It was previously thought that tumor cells switch expression of PKM from normal tissue-expressed PKM1 to tumor-specific PKM2 via an alternative splicing mechanism. This view was challenged by a recent report demonstrating that PKM2 is already the major PKM isoform expressed in many differentiated normal tissues. Here, through analyses on sixteen tumor types using the cancer genome atlas RNA-Seq and exon array datasets, we confirmed that isoform switch from PKM1 to PKM2 occurred in glioblastomas but not in other tumor types examined. Despite lacking of isoform switches, PKM2 expression was found to be increased in all cancer types examined, and correlated strongly to poor prognosis in head and neck cancers. We further demonstrated that elevated PKM2 expression correlated well with the hypomethylation status of intron 1 of the PKM gene in multiple cancer types, suggesting epigenetic regulation by DNA methylation as a major mechanism in controlling PKM transcription in tumors. Our study suggests that isoform switch of PKM1 to PKM2 in cancers is tissue-specific and targeting PKM2 activity in tumors remains a promising approach for clinical intervention of multiple cancer types.

Comparison of ultrasonic array imaging algorithms for nondestructive evaluation

Ultrasonic array imaging algorithms have been widely used and developed in nondestructive evaluation in the last 10 years. In this paper, three imaging algorithms [total focusing method (TFM), phase-coherent imaging (PCI), and spatial compounding imaging (SCI)] are compared through both simulation and experimental measurements. In the simulation, array data sets were generated using a hybrid forward model containing a single defect among a multitude of randomly distributed point scatterers to represent backscatter from material microstructure. The number of point scatterers per unit area and their scattering amplitude were optimized to reduce computation cost. The SNR of the final images and their resolution were used to indicate the quality of the different imaging algorithms. The images of different types of defects (point reflectors and planar cracks) were used to investigate the robustness of the imaging algorithms. It is shown that PCI can yield higher image resolution and higher SNR for defects in material with weak backscatter than TFM, but that the images of cracks are distorted. Overall, TFM is the most robust algorithm across a range of different types of defects. It is also shown that the detection limit of all three imaging algorithms is almost equal for weakly scattering defects.

Analysis of boutique arrays: A universal method for the selection of the optimal data normalization procedure

DNA microarrays, which are among the most popular genomic tools, are widely applied in biology and medicine. Boutique arrays, which are small, spotted, dedicated microarrays, constitute an inexpensive alternative to whole-genome screening methods. The data extracted from each microarray-based experiment must be transformed and processed prior to further analysis to eliminate any technical bias. The normalization of the data is the most crucial step of microarray data pre-processing and this process must be carefully considered as it has a profound effect on the results of the analysis. Several normalization algorithms have been developed and implemented in data analysis software packages. However, most of these methods were designed for whole-genome analysis. In this study, we tested 13 normalization strategies (ten for double-channel data and three for single-channel data) available on R Bioconductor and compared their effectiveness in the normalization of four boutique array datasets. The results revealed that boutique arrays can be successfully normalized using standard methods, but not every method is suitable for each dataset. We also suggest a universal seven-step workflow that can be applied for the selection of the optimal normalization procedure for any boutique array dataset. The described workflow enables the evaluation of the investigated normalization methods based on the bias and variance values for the control probes, a differential expression analysis and a receiver operating characteristic curve analysis. The analysis of each component results in a separate ranking of the normalization methods. A combination of the ranks obtained from all the normalization procedures facilitates the selection of the most appropriate normalization method for the studied dataset and determines which methods can be used interchangeably.

SAT0012 Estimating heritability of response to treatment with anti-TNF biologic agents using linear mixed models

BackgroundThe effectiveness of treatment with anti-TNF inhibitors in rheumatoid arthritis (RA) is variable with a significant minority of patients experiencing poor response. Ideally clinicians would like to identify which patients...

The origin of D″ reflections: a systematic study of seismic array data sets

Regional seismic studies of the Earth's deep mantle indicate the presence of P- and S-wave reflectors at the top of the D′′ layer. Such reflectors appear to be laterally intermittent, occur at a range of depths, and produce waveforms whose amplitude and polarity differ at different locations. A number of possible explanations for D′′ reflections have been proposed, including: a phase change from perovskite to post-perovskite; thermal or thermochemical discontinuities generated by subducted slabs; alignment of anisotropic minerals and global or small-scale chemical heterogeneity arising perhaps from core–mantle interactions or primitive material. However, some of the observed heterogeneity in D′′ structure may be related to lack of consistency between the style, quality and number of seismic observations in different regions, as well as complexity in wave propagation due to 3-D structure. In this study, we present a compilation of observations of D′′ reflections for regions in which (1) a reflector is seen in both P- and S-wave arrivals and (2) consistent array seismology techniques have been applied to infer the properties of the reflectors. Using a recent Monte Carlo thermodynamic modelling method to convert changes in chemistry and mineralogy into changes in seismic velocity, we compare the fit of different isotropic thermochemical structures to the observed seismic properties of D′′ reflections, namely the polarity and amplitude of the waveforms. Our results indicate that it may be possible to discriminate between different possible origins for D′′ reflections on the basis of polarity observations alone. Amplitude measurements may provide further constraints but must be treated with caution due to a large number of factors, which can influence waveform amplitudes. In regions such as the Caribbean, which are characterized by high-amplitude, negative polarity P waveforms and positive polarity S waveforms, a phase change from perovskite to post-perovskite provides the most likely explanation for the discontinuity. However, in other regions, notably Siberia and the central Pacific, P- and S waveforms both display a positive polarity, and multiple explanations for the discontinuity are possible, including subduction-related chemical heterogeneity and anisotropy.

Highlights from Recent Cancer Literature

Highlights from Recent Cancer Literature

Abstract 1667: Autophagy and unfolded protein response (UPR) signaling regulates progression of apoptosis in mammary gland involution.

Abstract Progression of local benign mammary lesions to invasive breast cancer is known to involve processes that are necessary for normal mammary gland development. Post-natal mammary gland is capable of undergoing repeated cycles of proliferation and cell death, most strikingly when fully differentiated (lactating) gland degenerates to a pre-lactation state during involution. Involution involves extensive cell death and tissue remodeling, which may temporarily increase breast cancer risk. Accumulation of milk proteins in the secretory epithelium creates the stress signal that triggers involution; however it is largely unknown whether the well-defined stress pathways, UPR and autophagy, are involved breast involution regulation. Using published gene expression array datasets from different phases of mouse mammary gland development we demonstrate that UPR, autophagy and apoptosis genes grouped in separate clusters. Next, in time-course experiments we show that autophagy and UPR signaling are tightly co-regulated during mammary involution. Early UPR signaling events such as GRP78 and phospho-eIF2α were elevated during early involution, i.e. the first 24 - 48 hours (h; reversible phase), while late UPR signaling events such as ATF4 and CHOP were upregulated during 72 h - 7 days of involution. An increase in AMPK was observed during early involution potentially preventing TORC1-mediated autophagy inhibition. Correlating with these observations, increased LC3-GFP punctate and decreased p62 expression were observed at 24 h when compared to 72 h post weaning, suggesting increased autophagy. Apoptosis was increased 48-72 h, peaking at 72 h involution, as measured by TUNEL. Increased apoptotic markers, such as cleaved caspase-7 and cleaved PARP were observed during these time points, with increased antiapoptotic BCL2 family members (BCL-XL and BCL-W) expressed from 24-48 h. To confirm the causal and inverse relationship between autophagy and apoptosis signaling, we performed involution time-course experiments using both low-dose drug interventions and an autophagy-related gene 7 (ATG7) deletion mouse model. Inhibition of autophagy by chloroquine, or genetic deletion of one ATG7 allele, enhanced the progression of mammary involution into an irreversible phase, characterized by earlier induction of apoptosis. In contrast, stimulation of autophagy by low dose tunicamycin treatment reduced apoptosis and extended the reversible phase of mammary involution by sustaining the secretory epithelium, suggesting a possible therapeutic use of autophagy stimulators in promoting lactation and nursing. Taken together these data indicate that UPR and autophagy play a key role in the regulation between survival and apoptosis during normal mammary gland involution and homeostasis, and, thus, if deregulated could contribute to breast cancer progression. Citation Format: Katherine L. Cook, Anni Warri, Rong Hu, Lu Jin, Alan Zwart, David R. Soto Pantoja, Jie Liu, Toren Finkel, Robert Clarke. Autophagy and unfolded protein response (UPR) signaling regulates progression of apoptosis in mammary gland involution. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 1667. doi:10.1158/1538-7445.AM2013-1667

Abstract 63: Incorporation of flanking probes reduces truncation losses for fluorescence in situ hybridization analysis of recurrent genomic deletions in tumor sections.

Abstract Fluorescence in situ hybridization (FISH) is a robust technique when used with appropriate interpretative guidelines, and the assay can yield consistent results that provide diagnostically and prognostically useful information to help guide patient therapy. The establishment of quality control standards in clinical laboratories using routine FISH analysis of formalin fixed paraffin embedded (FFPE) sections in different tumor types for fusion and break-apart strategies and gene amplifications have been developed for both analysis and reporting, but at the present time there are no comprehensive guidelines for the analysis of genomic deletions using FFPE sections. The use of FISH on archival FFPE samples is technically demanding and becomes more challenging when applied to paraffin-embedded tissue microarrays. The evaluation of FISH signals in interphase nuclei of FFPE sections is affected by truncation and overlapping of the nuclei due to varying cell density, tissue architecture differences and histological forms. In this study we report a generalizable four-color deletion FISH approach to assist interpretation problems arising when evaluating FISH signals. The guidelines will help address interpretative dilemmas associated with overlapping and truncated nuclei in FFPE prostate cancer sections. The four-color FISH approach was developed using the PTEN tumor suppressor gene deletion model in prostate cancer. The PTEN assay was based on a robust bioinformatics analysis of 311 published human genome array datasets and comprises a centromeric “chromosome enumeration” probe, a specific PTEN gene probe, and control flanking probes either side of the target probe. The sensitivity and specificity parameters of the four-color PTEN probe set were further characterized using a large number of well-characterized tumors and stringent scoring criteria. The incorporation of flanking control probes allowed the analysts to determine if the chromosomal region was subject to truncation loss. A minimum threshold for apparent deletion frequency was set to address the heterogeneous and homogeneous nature of tumor histology. In addition the approach facilitated analysis of genotypic heterogeneity and varying clonality within different foci of tumor in the prostate. Overall the approach provided robust and highly reproducible results that minimized inter- and intra-assay variability. The four-color FISH deletion assay reduced the frequency of misinterpretation and improved both the quality and throughput of FISH analyses using clinical samples. Moreover the use of established controls and conservative cut-offs for assigning deletions will facilitate more coherent approach to developing reporting standards for deletion assays as more tumor suppressor genes of clinical importance are discovered by next generation sequencing methods. Citation Format: Maisa Yoshimoto, Olga Ludkovski, Jennifer Good, Robert J. Gooding, Jean McGowan-Jordan, Alexander Boag, Andrew Evans, Ming-Sound Tsao, Paulo Nuin, Jeremy A. Squire. Incorporation of flanking probes reduces truncation losses for fluorescence in situ hybridization analysis of recurrent genomic deletions in tumor sections. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 63. doi:10.1158/1538-7445.AM2013-63

Abstract 3968: Ovarian cancer chemokine microenvironment is conducive to homing of CD3/CD28 co-stimulated T cells prepared for adoptive transfer therapy.

Abstract Chemokines play a central role in lymphocyte trafficking and therefore affect the success of immune therapy by supporting the homing of T cells to the site of disease. We are conducting a phase I study using CD3/CD28 co-stimulated lysate vaccine-primed autologous T cells for advanced ovarian cancer. Papillary serous ovarian cancer is a very heterogeneous tumor and little is known about the expression of chemokines in its microenvironment. We thus mapped the chemokines predominantly expressed in advanced stage papillary serous ovarian cancer and characterized the chemokine receptor expression and homing ability of co-stimulated autologous T-cells of patients participating in our adoptive T cell transfer trial. The expression of all known human chemokines was analyzed on a gene expression array dataset from 63 patients with primary ovarian cancer and validated on publicly available microarray dataset from 222 ovarian cancer samples, as well as on 17 primary ovary cancer lines. Immunohistochemistry for the 10 most highly expressed chemokines from the 3 databases was performed on tissue microarrays constructed from 50 patients with advanced stage papillary serous ovarian cancer. Slides were also stained for CD3+, CD8+ and FoxP3+ T cells. Median expression level of chemokines and T cells were measured at primary and metastatic sites. Vaccine-primed, ex vivo CD3/CD28 co-stimulated autologous T-cells of 5 patients were characterized for chemokine receptors by flow cytometry. Chemotaxis assays were performed to assess the functionality of the receptors. The 10 most highly expressed chemokines found in advanced stage papillary cancer by gene expression analysis were: CCL2, CCL4, CCL5, CCL28, CXCL5, CXCL6 CXCL10, CXCL12, CXCL16 and CX3CL1. Immunohistochemistry confirmed the expression of these chemokines at both primary and metastatic sites. The expression of CCL4 and CCL5 showed the strongest correlation with the presence of tumor-infiltrating CD8+ cells, whereas CCL28, CXCL12, CX3CL1, CXCL5 and CXCL6 expression showed correlation with increased numbers of FoxP3+ cells in tumor islets. The three expressed chemokine receptors on vaccine-primed CD3/CD28 co-stimulated T cells were CCR10 (CCL27, CCL28 receptor), CXCR3 (CXCL9-11 receptor) and CXCR4 (CXCL12 receptor). Chemotaxis assays confirmed the migration of these co-stimulated T cells towards ligands of all three of these receptors. Chemokines play a paramount role in tumor immune response and characterization of the tumor microenvironment is key to successful immunotherapy. Ex vivo expanded tumor specific cytotoxic T cells have shown promising results for the treatment of small-volume disease, and our data confirm that these cells are armed with the appropriate receptors to home to ovarian cancer. Citation Format: Emese Zsiros, Hongzhe Lee, Ian S. Hagemann, Priyanka Duttagupta, Renee Frank, Thomas Garrabrant, Bruce L. Levine, Carl H. June, Janos Tanyi, Lana Kandalaft, Michael Feldman, George Coukos. Ovarian cancer chemokine microenvironment is conducive to homing of CD3/CD28 co-stimulated T cells prepared for adoptive transfer therapy. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 3968. doi:10.1158/1538-7445.AM2013-3968

Abstract 230: Connectivity-map analyses identify FDA-approved drugs targeting breast cancer stem cells.

Abstract Recently developed genomics-based tools are allowing repositioning of established drugs as cancer treatments. This approach has not been employed extensively to identify drugs that may target cancer stem cells (CSCs) of solid tumors. CD44+/CD24- cancer cells are the first described CSCs for breast cancer, which are enriched in mesenchymal stem-like (MSL) and mesenchymal-like (ML) subtypes of triple negative (TNBCs) and in anti-estrogen resistant breast cancers. We subjected gene expression array datasets containing list of genes differentially expressed in CD44+/CD24- CSCs compared to non-CSCs as well as in cells enriched for other markers of CSCs including CD271 and GD3S (six datasets in total) to connectivity map (CMAP) to identify drugs that may reverse gene expression patterns unique to CSCs. All trans retinoic acid (ATRA) was negatively connected with gene expression in CD44+/CD24-, CD271+, and GD3S+ CSCs as well as in ML and MSL subtypes of TNBCs. Independently, ATRA was found to modulate the expression of several genes in the recently described CSC-enriched Yes-Associated Protein (YAP) signature. ATRA-inducible genes of the YAP signature were associated with good prognosis, whereas ATRA-repressible genes were associated with poor prognosis, depending on the subtype of breast cancer. ATRA reduced mammosphere forming ability of breast cancer cells including acquired anti-estrogen resistant cells, which correlated with ATRA-mediated reduction of the critical stemness-associated gene SOX2 and induction of SOX21, which functions as a differentiation factor depending on the cell type. CD44+/CD24- breast cancer cell line with K-ras mutation was resistant to ATRA, which was reversed by MAP kinase inhibitors. From these results, we propose ATRA as a potential therapy for specific subtypes of breast cancers enriched for CSCs. A biomarker-driven patient selection may help to identify patients who may benefit from ATRA therapy. Citation Format: Poornima Bhat-Nakshatri, Chirayu Goswami, Sunil Badve, George W. Sledge, Harikrishna Nakshatri. Connectivity-map analyses identify FDA-approved drugs targeting breast cancer stem cells. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 230. doi:10.1158/1538-7445.AM2013-230

Global Comparisons of Lectin–Glycan Interactions Using a Database of Analyzed Glycan Array Data

Lectin-glycan interactions have critical functions in multiple normal and pathological processes, but the binding partners and functions for many glycans and lectins are not known. An important step in better understanding glycan-lectin biology is enabling systematic quantification and analysis of the interactions. Glycan arrays can provide the experimental information for such analyses, and the thousands of glycan array datasets available through the Consortium for Functional Glycomics provide the opportunity to extend the analyses to a broad scale. We developed software, based on our previously described Motif Segregation algorithm, for the automated analysis of glycan array data, and we analyzed the entire storehouse of 2883 datasets from the Consortium for Functional Glycomics. We mined the resulting database to make comparisons of specificities across multiple lectins and comparisons between glycans in their lectin receptors. Of the lectins in the database, viral lectins were the most different from other organism types, with specificities nearly always restricted to sialic acids, and mammalian lectins had the most diverse range of specificities. Certain mammalian lectins were unique in their specificities for sulfated glycans. Simple modifications to a lactosamine core structure radically altered the types of lectins that were highly specific for the glycan. Unmodified lactosamine was specifically recognized by plant, fungal, viral, and mammalian lectins; sialylation shifted the binding mainly to viral lectins; and sulfation resulted in mainly mammalian lectins with the highest specificities. We anticipate that this analysis program and database will be valuable in fundamental glycobiology studies, detailed analyses of lectin specificities, and practical applications in translational research.

A scale-space method for detecting recurrent DNA copy number changes with analytical false discovery rate control

Tumor formation is partially driven by DNA copy number changes, which are typically measured using array comparative genomic hybridization, SNP arrays and DNA sequencing platforms. Many techniques are available for detecting recurring aberrations across multiple tumor samples, including CMAR, STAC, GISTIC and KC-SMART. GISTIC is widely used and detects both broad and focal (potentially overlapping) recurring events. However, GISTIC performs false discovery rate control on probes instead of events. Here we propose Analytical Multi-scale Identification of Recurrent Events, a multi-scale Gaussian smoothing approach, for the detection of both broad and focal (potentially overlapping) recurring copy number alterations. Importantly, false discovery rate control is performed analytically (no need for permutations) on events rather than probes. The method does not require segmentation or calling on the input dataset and therefore reduces the potential loss of information due to discretization. An important characteristic of the approach is that the error rate is controlled across all scales and that the algorithm outputs a single profile of significant events selected from the appropriate scales. We perform extensive simulations and showcase its utility on a glioblastoma SNP array dataset. Importantly, ADMIRE detects focal events that are missed by GISTIC, including two events involving known glioma tumor-suppressor genes: CDKN2C and NF1.

Estimation of a semiparametric mixture of regressions model

We introduce in this paper a new mixture of regressions model which is a generalisation of the semiparametric two-component mixture model studied in Bordes, Delmas, and Vandekerkhove [(2006b), ‘Semiparametric Estimation of a Two-component Mixture Model When a Component is Known’, Scandinavian Journal of Statistics, 33, 733–752]. Namely, we consider a two-component mixture of regressions model in which one component is entirely known while the proportion, the slope, the intercept, and the error distribution of the other component are unknown. Our model is said to be semiparametric in the sense that the probability density function (pdf) of the error involved in the unknown regression model cannot be modelled adequately by using a parametric density family. When the pdfs of the errors involved in each regression model are supposed to be zero-symmetric, we propose an estimator of the various (Euclidean and functional) parameters of the model, and establish under mild conditions their almost sure rates of convergence. Finally, the implementation and numerical performances of our method are discussed using several simulated data sets and one real high-density array data set (ChIP-mix model).

Elafin, an inhibitor of elastase, is a prognostic indicator in breast cancer

IntroductionElafin is an elastase-specific inhibitor with increased transcription in normal mammary epithelial cells compared to mammary carcinoma cells. In this report, we test the hypothesis that inhibition of elastase, through induction of elafin, leads to inhibition of human breast cancer cell viability and, therefore, predicts survival in breast cancer patients.MethodsPanels of normal and immortalized breast epithelial cells, along with breast carcinoma cells, were used to examine the impact of adenoviral-mediated elafin expression or shRNA-mediated inhibition of elastase on the growth of cells and xenografts in nude mice. To determine the prognostic significance of decreased elafin in patients with invasive breast cancer, previously published gene array datasets were interrogated.ResultsElafin expression had no effect on non-tumorigenic cells but resulted in marked inhibition of cell growth in breast cancer cell lines. Control-treated xenografts generated a tumor burden that necessitated sacrifice within one month of initial treatment, whereas xenograft-bearing mice treated with Ad-Elafin were alive at eight months with marked reduction in tumor growth. Elastase inhibition mimicked these results, showing decreased tumor cell growth in vitro and in vivo. Low expression of elafin gene correlated with significantly reduced time to relapse, and when combined with high expression of elastase gene was associated with decreased survival in breast cancer patients.ConclusionOur data suggest that elafin plays a direct role in the suppression of tumors through inhibition of elastase and thus serves as a prognostic indicator for breast cancer patients.

Comparison of Nasal Epithelial Smoking-Induced Gene Expression on Affymetrix Exon 1.0 and Gene 1.0 ST Arrays

We have previously defined the impact of tobacco smoking on nasal epithelium gene expression using Affymetrix Exon 1.0 ST arrays. In this paper, we compared the performance of the Affymetrix GeneChip Human Gene 1.0 ST array with the Human Exon 1.0 ST array for detecting nasal smoking-related gene expression changes. RNA collected from the nasal epithelium of five current smokers and five never smokers was hybridized to both arrays. While the intersample correlation within each array platform was relatively higher in the Gene array than that in the Exon array, the majority of the genes most changed by smoking were tightly correlated between platforms. Although neither array dataset was powered to detect differentially expressed genes (DEGs) at a false discovery rate (FDR) <0.05, we identified more DEGs than expected by chance using the Gene ST array. These findings suggest that while both platforms show a high degree of correlation for detecting smoking-induced differential gene expression changes, the Gene ST array may be a more cost-effective platform in a clinical setting for gene-level genomewide expression profiling and an effective tool for exploring the host response to cigarette smoking and other inhaled toxins.

Quality assessment and data handling methods for Affymetrix Gene 1.0 ST arrays with variable RNA integrity

BackgroundRNA and microarray quality assessment form an integral part of gene expression analysis and, although methods such as the RNA integrity number (RIN) algorithm reliably asses RNA integrity, the relevance of RNA integrity in gene expression analysis as well as analysis methods to accommodate the possible effects of degradation requires further investigation. We investigated the relationship between RNA integrity and array quality on the commonly used Affymetrix Gene 1.0 ST array platform using reliable within-array and between-array quality assessment measures. The possibility of a transcript specific bias in the apparent effect of RNA degradation on the measured gene expression signal was evaluated after either excluding quality-flagged arrays or compensation for RNA degradation at different steps in the analysis.ResultsUsing probe-level and inter-array quality metrics to assess 34 Gene 1.0 ST array datasets derived from historical, paired tumour and normal primary colorectal cancer samples, 7 arrays (20.6%), with a mean sample RIN of 3.2 (SD = 0.42), were flagged during array quality assessment while 10 arrays from samples with RINs < 7 passed quality assessment, including one sample with a RIN < 3. We detected a transcript length bias in RNA degradation in only 5.8% of annotated transcript clusters (p-value 0.05, FC ≥ |2|), with longer and shorter than average transcripts under- and overrepresented in quality-flagged samples respectively. Applying compensatory measures for RNA degradation performed at least as well as excluding quality-flagged arrays, as judged by hierarchical clustering, gene expression analysis and Ingenuity Pathway Analysis; importantly, use of these compensatory measures had the significant benefit of enabling lower quality array data from irreplaceable clinical samples to be retained in downstream analyses.ConclusionsHere, we demonstrate an effective array-quality assessment strategy, which will allow the user to recognize lower quality arrays that can be included in the analysis once appropriate measures are applied to account for known or unknown sources of variation, such as array quality- and batch- effects, by implementing ComBat or Surrogate Variable Analysis. This approach of quality control and analysis will be especially useful for clinical samples with variable and low RNA qualities, with RIN scores ≥ 2.

DNA methylation age of human tissues and cell types

BackgroundIt is not yet known whether DNA methylation levels can be used to accurately predict age across a broad spectrum of human tissues and cell types, nor whether the resulting age prediction is a biologically meaningful measure.ResultsI developed a multi-tissue predictor of age that allows one to estimate the DNA methylation age of most tissues and cell types. The predictor, which is freely available, was developed using 8,000 samples from 82 Illumina DNA methylation array datasets, encompassing 51 healthy tissues and cell types. I found that DNA methylation age has the following properties: first, it is close to zero for embryonic and induced pluripotent stem cells; second, it correlates with cell passage number; third, it gives rise to a highly heritable measure of age acceleration; and, fourth, it is applicable to chimpanzee tissues. Analysis of 6,000 cancer samples from 32 datasets showed that all of the considered 20 cancer types exhibit significant age acceleration, with an average of 36 years. Low age-acceleration of cancer tissue is associated with a high number of somatic mutations and TP53 mutations, while mutations in steroid receptors greatly accelerate DNA methylation age in breast cancer. Finally, I characterize the 353 CpG sites that together form an aging clock in terms of chromatin states and tissue variance.ConclusionsI propose that DNA methylation age measures the cumulative effect of an epigenetic maintenance system. This novel epigenetic clock can be used to address a host of questions in developmental biology, cancer and aging research.