Stochastic segmentation models for array-based comparative genomic hybridization data analysis

Array-based comparative genomic hybridization (array-CGH) is a high throughput, high resolution technique for studying the genetics of cancer. Analysis of array-CGH data typically involves estimation of the underlying chromosome copy numbers from the log fluorescence ratios and segmenting the chromosome into regions with the same copy number at each location. We propose for the analysis of array-CGH data, a new stochastic segmentation model and an associated estimation procedure that has attractive statistical and computational properties. An important benefit of this Bayesian segmentation model is that it yields explicit formulas for posterior means, which can be used to estimate the signal directly without performing segmentation. Other quantities relating to the posterior distribution that are useful for providing confidence assessments of any given segmentation can also be estimated by using our method. We propose an approximation method whose computation time is linear in sequence length which makes our method practically applicable to the new higher density arrays. Simulation studies and applications to real array-CGH data illustrate the advantages of the proposed approach.     

Differential gene alteration among hepatoma cell lines demonstrated by cDNA microarray-based comparative genomic hybridization

We assayed chromosomal abnormalities in hepatoma cell lines using the microarray-based comparative genomic hybridization (array-CGH) method and investigated the relationship between genomic copy number alterations and expression profiles in these hepatoma cell lines. We modified a cDNA array-CGH assay to compare genomic DNAs from seven hepatoma cell lines, as well as DNA from two non-hepatoma cell lines and from normal cells. The mRNA expression of each sample was assayed in parallel by cDNA microarray. We identified small amplified or deleted chromosomal regions, as well as alterations in DNA copy number not previously described. We predominantly found alterations of apoptosis-related genes in Hep3B and HepG2, cell adhesion and receptor molecules in HLE, and cytokine-related genes in PLC/PRF/5. About 40% of the genes showing amplification or loss showed altered levels of mRNA (p < 0.05). Hierarchical clustering analysis showed that the expression of these genes allows differentiation between alpha-fetoprotein (AFP)-producing and AFP-negative cell lines. cDNA array-CGH is a sensitive method that can be used to detect alterations in genomic copy number in tumor cells. Differences in DNA copy alterations between AFP-producing and AFP-negative cells may lead to differential gene expression and may be related to the phenotype of these cells.     

High resolution array-CGH analysis of single cells

Heterogeneity in the genome copy number of tissues is of particular importance in solid tumor biology. Furthermore, many clinical applications such as pre-implantation and non-invasive prenatal diagnosis would benefit from the ability to characterize individual single cells. As the amount of DNA from single cells is so small, several PCR protocols have been developed in an attempt to achieve unbiased amplification. Many of these approaches are suitable for subsequent cytogenetic analyses using conventional methodologies such as comparative genomic hybridization (CGH) to metaphase spreads. However, attempts to harness array-CGH for single-cell analysis to provide improved resolution have been disappointing. Here we describe a strategy that combines single-cell amplification using GenomePlex library technology (GenomePlex^® Single Cell Whole Genome Amplification Kit, Sigma-Aldrich, UK) and detailed analysis of genomic copy number changes by high-resolution array-CGH. We show that single copy changes as small as 8.3 Mb in single cells are detected reliably with single cells derived from various tumor cell lines as well as patients presenting with trisomy 21 and Prader–Willi syndrome. Our results demonstrate the potential of this technology for studies of tumor biology and for clinical diagnostics.     

Spatial normalization of array-CGH data

Background  

Array-based comparative genomic hybridization (array-CGH) is a recently developed technique for analyzing changes in DNA copy number. As in all microarray analyses, normalization is required to correct for experimental artifacts while preserving the true biological signal. We investigated various sources of systematic variation in array-CGH data and identified two distinct types of spatial effect of no biological relevance as the predominant experimental artifacts: continuous spatial gradients and local spatial bias. Local spatial bias affects a large proportion of arrays, and has not previously been considered in array-CGH experiments.     

Analysis of array CGH data for cancer studies using fused quantile regression

MOTIVATION: The identification of DNA copy number changes provides insights that may advance our understanding of initiation and progression of cancer. Array-based comparative genomic hybridization (array-CGH) has emerged as a technique allowing high-throughput genome-wide scanning for chromosomal aberrations. A number of statistical methods have been proposed for the analysis of array-CGH data. In this article, we consider a fused quantile regression model based on three motivations: (1) quantile regression may provide a more comprehensive picture for the ratio profile of copy numbers than the standard mean regression approach; (2) for simplicity, most available methods assume uniform spacing between neighboring clones, while incorporating the information of physical locations of clones may be helpful and (3) most current methods have a set of tuning parameters that must be carefully tuned, which introduces complexity to the implementation. RESULTS: We formulate the detection of regions of gains and losses in a fused regularized quantile regression framework, incorporating physical locations of clones. We derive an efficient algorithm that computes the entire solution path for the resulting optimization problem, and we propose a simple estimate for the complexity of the fitted model, which leads to convenient selection of the tuning parameter. Three published array-CGH datasets are used to demonstrate our approach. AVAILABILITY: R code are available at http://www.stat.lsa.umich.edu/~jizhu/code/cgh/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.     

ArrayCyGHt: a web application for analysis and visualization of array-CGH data

ArrayCyGHt is a web-based application tool for analysis and visualization of microarray-comparative genomic hybridization (array-CGH) data. Full process of array-CGH data analysis, from normalization of raw data to the final visualization of copy number gain or loss, can be straightforwardly achieved on this arrayCyGHt system without the use of any further software. ArrayCyGHt, therefore, provides an easy and fast tool for the analysis of copy number aberrations in any kinds of data format. AVAILABILITY: ArrayCyGHt can be accessed at http://genomics.catholic.ac.kr/arrayCGH/     

Genomic microarrays in the spotlight

Microarray-based comparative genomic hybridization (array-CGH) has emerged as a revolutionary platform, enabling the high-resolution detection of DNA copy number aberrations. In this article we outline the use and limitations of genomic clones, cDNA clones and PCR products as targets for genomic microarray construction. Furthermore, the applications and future aspects of these arrays for DNA copy number analysis in research and diagnostics, epigenetic profiling and gene annotation are discussed. These recent developments of genomic microarrays mark only the beginning of a new generation of high-resolution and high-throughput tools for genetic analysis.     

Computation of recurrent minimal genomic alterations from array-CGH data

MOTIVATION: The identification of recurrent genomic alterations can provide insight into the initiation and progression of genetic diseases, such as cancer. Array-CGH can identify chromosomal regions that have been gained or lost, with a resolution of approximately 1 mb, for the cutting-edge techniques. The extraction of discrete profiles from raw array-CGH data has been studied extensively, but subsequent steps in the analysis require flexible, efficient algorithms, particularly if the number of available profiles exceeds a few tens or the number of array probes exceeds a few thousands. RESULTS: We propose two algorithms for computing minimal and minimal constrained regions of gain and loss from discretized CGH profiles. The second of these algorithms can handle additional constraints describing relevant regions of copy number change. We have validated these algorithms on two public array-CGH datasets. AVAILABILITY: From the authors, upon request. CONTACT: celine@lri.fr SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.     

CGcgh: a tool for molecular karyotyping using DNA microarray-based comparative genomic hybridization (array-CGH)

Microarray-based comparative genomic hybridization (array-CGH) is a technique by which variations in copy numbers between two genomes can be analyzed using DNA microarrays. Array CGH has been used to survey chromosomal amplifications and deletions in fetal aneuploidies or cancer tissues. Herein we report a user-friendly, MATLAB-based, array CGH analyzing program, Chang Gung comparative genomic hybridization (CGcgh), as a standalone PC version. The analyzed chromosomal data are displayed in a graphic interface, and CGcgh allows users to launch a corresponding G-banding ideogram. The abnormal DNA copy numbers (gains and losses) can be identified automatically using a user defined window size (default value is 50 probes) and sequential student t-tests with sliding windows along with chromosomes. CGcgh has been tested in multiple karyotype-confirmed human samples, including five published cases and trisomies 13, 18, 21 and X from our laboratories, and 18 cases of which microarray data are available publicly. CGcgh can be used to detect the copy number changes in small genomic regions, which are commonly encountered by clinical geneticists. CGcgh works well for the data from cDNA microarray, spotted oligonucleotide microarrays, and Affymetrix Human Mapping Arrays (10K, 100K, 500K Array Sets). The program can be freely downloaded from . Y. S. Lee and A. Chao contributed equally to this work.     

Genome-wide profiling of gene amplification and deletion in cancer

Accumulations of genetic changes in somatic cells induce phenotypic transformations leading to cancer. Among these genetic changes, gene amplification and deletion are most frequently observed in several kinds of cancers. Amplification of oncogene and/or deletion of tumor suppressor gene, together with dysfunction of the gene by point mutation, are the main causes of cancer. Genome-wide analysis of amplification and deletion of genes in cancers is basic to resolving the mechanisms of carcinogenesis. Comparative genomic hybridization (CGH) developed in 1992 has been utilized to identify DNA copy number abnormalities in various kind of cancers and several reports have shown its usefulness in screening of the genes involved in carcinogenesis, and also in the identification of prognostic factors in cancer. We have shown that 1q23 gain is associated with neuroblastomas that are resistant to aggressive treatment, and have poor prognosis, and 1q and 13q gains are possibly related to drug resistance in ovarian cancers. Recently, the "rough draft" of the human genome was reported and we are ready to utilize the vast information on genomic sequences in cancer research. Moreover, microarray technology enables us to analyze more than ten thousand genes at a time and revealed genetic abnormalities in cancers at a genome-wide level. By combination of microarray and CGH, a powerful screening method for oncogenes and tumor suppressor genes in cancers, called array-CGH, has been developed by several groups. In this article, we overview these genome-wide analytical methods, CGH and array-CGH, and discuss their potential in molecular characterization of cancers.     

Variation of CNV distribution in five different ethnic populations

Recent studies have revealed a new type of variation in the human genome encompassing relatively large genomic segments ( approximately 100 kb-2.5 Mb), commonly referred to as copy number variation (CNV). The full nature and extent of CNV and its frequency in different ethnic populations is still largely unknown. In this study we surveyed a set of 12 CNVs previously detected by array-CGH. More than 300 individuals from five different ethnic populations, including three distinct European, one Asian and one African population, were tested for the occurrence of CNV using multiplex ligation-dependent probe amplification (MLPA). Seven of these loci indeed showed CNV, i.e., showed copy numbers that deviated from the population median. More precise estimations of the actual genomic copy numbers for (part of) the NSF gene locus, revealed copy numbers ranging from two to at least seven. Additionally, significant inter-population differences in the distribution of these copy numbers were observed. These data suggest that insight into absolute DNA copy numbers for loci exhibiting CNV is required to determine their potential contribution to normal phenotypic variation and, in addition, disease susceptibility.     

SiDCoN: a tool to aid scoring of DNA copy number changes in SNP chip data

The recent application of genome-wide, single nucleotide polymorphism (SNP) microarrays to investigate DNA copy number aberrations in cancer has provided unparalleled sensitivity for identifying genomic changes. In some instances the complexity of these changes makes them difficult to interpret, particularly when tumour samples are contaminated with normal (stromal) tissue. Current automated scoring algorithms require considerable manual data checking and correction, especially when assessing uncultured tumour specimens. To address these limitations we have developed a visual tool to aid in the analysis of DNA copy number data. Simulated DNA Copy Number (SiDCoN) is a spreadsheet-based application designed to simulate the appearance of B-allele and logR plots for all known types of tumour DNA copy number changes, in the presence or absence of stromal contamination. The system allows the user to determine the level of stromal contamination, as well as specify up to 3 different DNA copy number aberrations for up to 5000 data points (representing individual SNPs). This allows users great flexibility to assess simple or complex DNA copy number combinations. We demonstrate how this utility can be used to estimate the level of stromal contamination within tumour samples and its application in deciphering the complex heterogeneous copy number changes we have observed in a series of tumours. We believe this tool will prove useful to others working in the area, both as a training tool, and to aid in the interpretation of complex copy number changes.     

Quantitative high-resolution genomic analysis of single cancer cells

During cancer progression, specific genomic aberrations arise that can determine the scope of the disease and can be used as predictive or prognostic markers. The detection of specific gene amplifications or deletions in single blood-borne or disseminated tumour cells that may give rise to the development of metastases is of great clinical interest but technically challenging. In this study, we present a method for quantitative high-resolution genomic analysis of single cells. Cells were isolated under permanent microscopic control followed by high-fidelity whole genome amplification and subsequent analyses by fine tiling array-CGH and qPCR. The assay was applied to single breast cancer cells to analyze the chromosomal region centred by the therapeutical relevant EGFR gene. This method allows precise quantitative analysis of copy number variations in single cell diagnostics.     

Weighted clustering of called array CGH data

Array comparative genomic hybridization (aCGH) is a laboratory technique to measure chromosomal copy number changes. A clear biological interpretation of the measurements is obtained by mapping these onto an ordinal scale with categories loss/normal/gain of a copy. The pattern of gains and losses harbors a level of tumor specificity. Here, we present WECCA (weighted clustering of called aCGH data), a method for weighted clustering of samples on the basis of the ordinal aCGH data. Two similarities to be used in the clustering and particularly suited for ordinal data are proposed, which are generalized to deal with weighted observations. In addition, a new form of linkage, especially suited for ordinal data, is introduced. In a simulation study, we show that the proposed cluster method is competitive to clustering using the continuous data. We illustrate WECCA using an application to a breast cancer data set, where WECCA finds a clustering that relates better with survival than the original one.     

Accurate, high-throughput typing of copy number variation using paralogue ratios from dispersed repeats

Recent work has demonstrated an unexpected prevalence of copy number variation in the human genome, and has highlighted the part this variation may play in predisposition to common phenotypes. Some important genes vary in number over a high range (e.g. DEFB4, which commonly varies between two and seven copies), and have posed formidable technical challenges for accurate copy number typing, so that there are no simple, cheap, high-throughput approaches suitable for large-scale screening. We have developed a simple comparative PCR method based on dispersed repeat sequences, using a single pair of precisely designed primers to amplify products simultaneously from both test and reference loci, which are subsequently distinguished and quantified via internal sequence differences. We have validated the method for the measurement of copy number at DEFB4 by comparison of results from >800 DNA samples with copy number measurements by MAPH/REDVR, MLPA and array-CGH. The new Paralogue Ratio Test (PRT) method can require as little as 10 ng genomic DNA, appears to be comparable in accuracy to the other methods, and for the first time provides a rapid, simple and inexpensive method for copy number analysis, suitable for application to typing thousands of samples in large case-control association studies.     

Chromosomal aberrations in bladder cancer: fresh versus formalin fixed paraffin embedded tissue and targeted FISH versus wide microarray-based CGH analysis

Bladder carcinogenesis is believed to follow two alternative pathways driven by the loss of chromosome 9 and the gain of chromosome 7, albeit other nonrandom copy number alterations (CNAs) were identified. However, confirmation studies are needed since many aspects of this model remain unclear and considerable heterogeneity among cases has emerged. One of the purposes of this study was to evaluate the performance of a targeted test (UroVysion assay) widely used for the detection of Transitional Cell Carcinoma (TCC) of the bladder, in two different types of material derived from the same tumor. We compared the results of UroVysion test performed on Freshly Isolated interphasic Nuclei (FIN) and on Formalin Fixed Paraffin Embedded (FFPE) tissues from 22 TCCs and we didn't find substantial differences. A second goal was to assess the concordance between array-CGH profiles and the targeted chromosomal profiles of UroVysion assay on an additional set of 10 TCCs, in order to evaluate whether UroVysion is an adequately sensitive method for the identification of selected aneuploidies and nonrandom CNAs in TCCs. Our results confirmed the importance of global genomic screening methods, that is array based CGH, to comprehensively determine the genomic profiles of large series of TCCs tumors. However, this technique has yet some limitations, such as not being able to detect low level mosaicism, or not detecting any change in the number of copies for a kind of compensatory effect due to the presence of high cellular heterogeneity. Thus, it is still advisable to use complementary techniques such as array-CGH and FISH, as the former is able to detect alterations at the genome level not excluding any chromosome, but the latter is able to maintain the individual data at the level of single cells, even if it focuses on few genomic regions.     

Optimized amplification and fluorescent labeling of small cell samples for genomic array-CGH.

BACKGROUND: Whole genome amplification (WGA) is usually needed in the genetic analysis of samples containing a low number of cells. In genome-wide analysis of DNA copy numbers by array comparative genomic hybridization (array-CGH) it is very important that the genome is evenly represented throughout the amplified product. All currently available WGA techniques are generating some degree of bias. METHODS: A way to compensate for this is using a reference sample which is similarly amplified, as the introduced amplification bias will be leveled out. Additionally, direct labeling of the amplified DNA is performed to bypass the currently widely applied random primed labeling, which involves an additional amplification of the product and is introducing extra bias. RESULTS: In this article it is shown that equal processing of the test and reference sample is indeed crucial to generate an optimal array-CGH profile of amplified DNA samples. Also presented here is that the labeling method may significantly effect the array-CGH result, it is shown that with direct chemical labeling using platinum derivates (ULS labeling) optimal array-CGH results are obtained. CONCLUSIONS: We show that an optimized WGA strategy for both test and reference sample in combination with direct chemical labeling results in a reliable array-CGH profile of samples as low as a 30 cell equivalent.     

Detection of gene copy number changes in CGH microarrays using a spatially correlated mixture model

MOTIVATION: Comparative genomic hybridization array experiments that investigate gene copy number changes present new challenges for statistical analysis and call for methods that incorporate spatial dependence between sequences along the chromosome. For this purpose, we propose a novel method called CGHmix. It is based on a spatially structured mixture model with three states corresponding to genomic sequences that are either unmodified, deleted or amplified. Inference is performed in a Bayesian framework. From the output, posterior probabilities of belonging to each of the three states are estimated for each genomic sequence and used to classify them. RESULTS: Using simulated data, CGHmix is validated and compared with both a conventional unstructured mixture model and with a recently proposed data mining method. We demonstrate the good performance of CGHmix for classifying copy number changes. In addition, the method provides a good estimate of the false discovery rate. We also present the analysis of a cancer related dataset. SUPPLEMENTARY INFORMATION: http://www.bgx.org.uk/papers.html     

Probe-free allele-specific copy number detection and analysis of tumors

Cancer development and progression frequently involve nucleotide mutations as well as amplifications and deletions of genomic segments. Quantification of allele-specific copy number is an important step in characterizing tumor genomes for precision medicine. Despite advances in approaches to high-throughput genomic DNA analysis, inexpensive and simple methods for analyzing complex nucleotide and copy number variants are still needed. Real-time polymerase chain reaction (PCR) methods for discovering and genotyping single nucleotide polymorphisms are becoming increasingly important in genetic analysis. In this study, we describe a simple, single-tube, probe-free method that combines SYBR Green I-based quantitative real-time PCR and quantitative melting curve analysis both to detect specific nucleotide variants and to quantify allele-specific copy number variants of tumors. The approach is based on the quantification of the targets of interest and the relative abundance of two alleles in a single tube. The specificity, sensitivity, and utility of the assay were demonstrated in detecting allele-specific copy number changes critical for carcinogenesis and therapeutic intervention. Our approach would be useful for allele-specific copy number analysis or precise genotyping.     

A segmental maximum a posteriori approach to genome-wide copy number profiling

MOTIVATION: Copy number profiling methods aim at assigning DNA copy numbers to chromosomal regions using measurements from microarray-based comparative genomic hybridizations. Among the proposed methods to this end, Hidden Markov Model (HMM)-based approaches seem promising since DNA copy number transitions are naturally captured in the model. Current discrete-index HMM-based approaches do not, however, take into account heterogeneous information regarding the genomic overlap between clones. Moreover, the majority of existing methods are restricted to chromosome-wise analysis. RESULTS: We introduce a novel Segmental Maximum A Posteriori approach, SMAP, for DNA copy number profiling. Our method is based on discrete-index Hidden Markov Modeling and incorporates genomic distance and overlap between clones. We exploit a priori information through user-controllable parameterization that enables the identification of copy number deviations of various lengths and amplitudes. The model parameters may be inferred at a genome-wide scale to avoid overfitting of model parameters often resulting from chromosome-wise model inference. We report superior performances of SMAP on synthetic data when compared with two recent methods. When applied on our new experimental data, SMAP readily recognizes already known genetic aberrations including both large-scale regions with aberrant DNA copy number and changes affecting only single features on the array. We highlight the differences between the prediction of SMAP and the compared methods and show that SMAP accurately determines copy number changes and benefits from overlap consideration.