Comparative analysis of algorithms for identifying amplifications and deletions in array CGH data

Asterias: a parallelized web-based suite for the analysis of expression and aCGH data

The analysis of expression and CGH arrays plays a central role in the study of complex diseases, especially cancer, including finding markers for early diagnosis and prognosis, choosing an optimal therapy, or increasing our understanding of cancer development and metastasis. Asterias (http://www.asterias.info) is an integrated collection of freely-accessible web tools for the analysis of gene expression and aCGH data. Most of the tools use parallel computing (via MPI) and run on a server with 60 CPUs for computation; compared to a desktop or server-based but not parallelized application, parallelization provides speed ups of factors up to 50. Most of our applications allow the user to obtain additional information for user-selected genes (chromosomal location, PubMed ids, Gene Ontology terms, etc.) by using clickable links in tables and/or figures. Our tools include: normalization of expression and aCGH data (DNMAD); converting between different types of gene/clone and protein identifiers (IDconverter/IDClight); filtering and imputation (preP); finding differentially expressed genes related to patient class and survival data (Pomelo II); searching for models of class prediction (Tnasas); using random forests to search for minimal models for class prediction or for large subsets of genes with predictive capacity (GeneSrF); searching for molecular signatures and predictive genes with survival data (SignS); detecting regions of genomic DNA gain or loss (ADaCGH). The capability to send results between different applications, access to additional functional information, and parallelized computation make our suite unique and exploit features only available to web-based applications.

Exploiting noise in array CGH data to improve detection of DNA copy number change

Developing effective methods for analyzing array-CGH data to detect chromosomal aberrations is very important for the diagnosis of pathogenesis of cancer and other diseases. Current analysis methods, being largely based on smoothing and/or segmentation, are not quite capable of detecting both the aberration regions and the boundary break points very accurately. Furthermore, when evaluating the accuracy of an algorithm for analyzing array-CGH data, it is commonly assumed that noise in the data follows normal distribution. A fundamental question is whether noise in array-CGH is indeed Gaussian, and if not, can one exploit the characteristics of noise to develop novel analysis methods that are capable of detecting accurately the aberration regions as well as the boundary break points simultaneously? By analyzing bacterial artificial chromosomes (BACs) arrays with an average 1 mb resolution, 19 k oligo arrays with the average probe spacing <100 kb and 385 k oligo arrays with the average probe spacing of about 6 kb, we show that when there are aberrations, noise in all three types of arrays is highly non-Gaussian and possesses long-range spatial correlations, and that such noise leads to worse performance of existing methods for detecting aberrations in array-CGH than the Gaussian noise case. We further develop a novel method, which has optimally exploited the character of the noise, and is capable of identifying both aberration regions as well as the boundary break points very accurately. Finally, we propose a new concept, posteriori signal-to-noise ratio (p-SNR), to assign certain confidence level to an aberration region and boundaries detected.

High-resolution copy number analysis of paraffin-embedded archival tissue using SNP BeadArrays

High-density SNP microarrays provide insight into the genomic events that occur in diseases like cancer through their capability to measure both LOH and genomic copy numbers. Where currently available methods are restricted to the use of fresh frozen tissue, we now describe the design and validation of copy number measurements using the Illumina BeadArray platform and the application of this technique to formalin-fixed, paraffin-embedded (FFPE) tissue. In fresh frozen tissue from a set of colorectal tumors with numerous chromosomal aberrations, our method measures copy number patterns that are comparable to values from established platforms, like Affymetrix GeneChip and BAC array-CGH. Moreover, paired comparisons of fresh frozen and FFPE tissues showed nearly identical patterns of genomic change. We conclude that this method enables the use of paraffin-embedded material for research into both LOH and numerical chromosomal abnormalities. These findings make the large pathological archives available for genomic analysis, which could be especially relevant for hereditary disease where fresh material from affected relatives is rarely available.

A faster circular binary segmentation algorithm for the analysis of array CGH data

MOTIVATION

Array CGH technologies enable the simultaneous measurement of DNA copy number for thousands of sites on a genome. We developed the circular binary segmentation (CBS) algorithm to divide the genome into regions of equal copy number. The algorithm tests for change-points using a maximal t-statistic with a permutation reference distribution to obtain the corresponding P-value. The number of computations required for the maximal test statistic is O(N2), where N is the number of markers. This makes the full permutation approach computationally prohibitive for the newer arrays that contain tens of thousands markers and highlights the need for a faster algorithm.

RESULTS

We present a hybrid approach to obtain the P-value of the test statistic in linear time. We also introduce a rule for stopping early when there is strong evidence for the presence of a change. We show through simulations that the hybrid approach provides a substantial gain in speed with only a negligible loss in accuracy and that the stopping rule further increases speed. We also present the analyses of array CGH data from breast cancer cell lines to show the impact of the new approaches on the analysis of real data.

AVAILABILITY

An R version of the CBS algorithm has been implemented in the "DNAcopy" package of the Bioconductor project. The proposed hybrid method for the P-value is available in version 1.2.1 or higher and the stopping rule for declaring a change early is available in version 1.5.1 or higher.

Asterias: A Parallelized Web-based Suite for the Analysis of Expression and aCGH Data

The analysis of expression and CGH arrays plays a central role in the study of complex diseases, especially cancer, including finding markers for early diagnosis and prognosis, choosing an optimal therapy, or increasing our understanding of cancer development and metastasis. Asterias ( http://www.asterias.info ) is an integrated collection of freely-accessible web tools for the analysis of gene expression and aCGH data. Most of the tools use parallel computing (via MPI) and run on a server with 60 CPUs for computation; compared to a desktop or server-based but not parallelized application, parallelization provides speed ups of factors up to 50. Most of our applications allow the user to obtain additional information for user-selected genes (chromosomal location, PubMed ids, Gene Ontology terms, etc.) by using clickable links in tables and/or figures. Our tools include: normalization of expression and aCGH data (DNMAD); converting between different types of gene/clone and protein identifiers (IDconverter/IDClight); filtering and imputation (preP); finding differentially expressed genes related to patient class and survival data (Pomelo II); searching for models of class prediction (Tnasas); using random forests to search for minimal models for class prediction or for large subsets of genes with predictive capacity (GeneSrF); searching for molecular signatures and predictive genes with survival data (SignS); detecting regions of genomic DNA gain or loss (ADaCGH). The capability to send results between different applications, access to additional functional information, and parallelized computation make our suite unique and exploit features only available to web-based applications.

Integrated high-resolution genome-wide analysis of gene dosage and gene expression in human brain tumors

A hallmark genomic feature of human brain tumors is the presence of multiple complex structural and numerical chromosomal aberrations that result in altered gene dosages. These genetic alterations lead to widespread, genome-wide gene expression changes. Both gene expression as well as gene copy number profiles can be assessed on a large scale using microarray methodology. The integration of genetic data with gene expression data provides a particularly effective approach for cancer gene discovery. Utilizing an array of bioinformatics tools, we describe an analysis algorithm that allows for the integration of gene copy number and gene expression profiles as a first-pass means of identifying potential cancer gene targets in human (brain) tumors. This strategy combines circular binary segmentation for the identification of gene copy number alterations, and gene copy number and gene expression data integration with a modification of signal-to-noise ratio computation and random permutation testing. We have evaluated this approach and confirmed its efficacy in the human glioma genome.

Characterizing genetically stable and unstable gastric cancers by microsatellites and array comparative genomic hybridization

Gastric cancer (GCA) displays a variety of genomic aberrations, including DNA copy number alterations, microsatellite instability (MSI), and loss of heterozygosity (LOH). The main aim of the present work was to determine the copy number aberrations in tumors with and without MSI or LOH. Fifteen fresh-frozen GCA samples, 11 of the intestinal and 4 of the diffuse type, were grouped by microsatellite analysis into high-level MSI (MSI-H, n = 2), LOH (n = 5), and microsatellite stable, LOH not detected (MSS/LOH-N, n = 8) tumors. The DNA samples were subsequently analyzed by array comparative genomic hybridization with 16,000 cDNA clones. As expected, the LOH tumors showed more copy number changes; however, the frequency of small-size amplifications was similar across all tumor groups. In addition, the cDNA arrays detected two apparently single-gene amplicons, at 11q13 (CCND1) and 12p12.1 (K-RAS), the presence of which were confirmed using oligonucleotide arrays. A novel amplicon at 5q13.2 was found only in diffuse-type tumors, which were otherwise genetically stable. The results suggest that DNA copy number changes may also occur in gastric cancers that show genomic stability in microsatellite analysis.

A genome-wide screen reveals functional gene clusters in the cancer genome and identifies EphA2 as a mitogen in glioblastoma

A novel genome-wide screen that combines patient outcome analysis with array comparative genomic hybridization and mRNA expression profiling was developed to identify genes with copy number alterations, aberrant mRNA expression, and relevance to survival in glioblastoma. The method led to the discovery of physical gene clusters within the cancer genome with boundaries defined by physical proximity, correlated mRNA expression patterns, and survival relatedness. These boundaries delineate a novel genomic interval called the functional common region (FCR). Many FCRs contained genes of high biological relevance to cancer and were used to pinpoint functionally significant DNA alterations that were too small or infrequent to be reliably identified using standard algorithms. One such FCR contained the EphA2 receptor tyrosine kinase. Validation experiments showed that EphA2 mRNA overexpression correlated inversely with patient survival in a panel of 21 glioblastomas, and ligand-mediated EphA2 receptor activation increased glioblastoma proliferation and tumor growth via a mitogen-activated protein kinase-dependent pathway. This novel genome-wide approach greatly expanded the list of target genes in glioblastoma and represents a powerful new strategy to identify the upstream determinants of tumor phenotype in a range of human cancers.

Comparison of chromosomal and array-based comparative genomic hybridization for the detection of genomic imbalances in primary prostate carcinomas

BACKGROUND

In order to gain new insights into the molecular mechanisms involved in prostate cancer, we performed array-based comparative genomic hybridization (aCGH) on a series of 46 primary prostate carcinomas using a 1 Mbp whole-genome coverage platform. As chromosomal comparative genomic hybridization (cCGH) data was available for these samples, we compared the sensitivity and overall concordance of the two methodologies, and used the combined information to infer the best of three different aCGH scoring approaches.

RESULTS

Our data demonstrate that the reliability of aCGH in the analysis of primary prostate carcinomas depends to some extent on the scoring approach used, with the breakpoint estimation method being the most sensitive and reliable. The pattern of copy number changes detected by aCGH was concordant with that of cCGH, but the higher resolution technique detected 2.7 times more aberrations and 15.2% more carcinomas with genomic imbalances. We additionally show that several aberrations were consistently overlooked using cCGH, such as small deletions at 5q, 6q, 12p, and 17p. The latter were validated by fluorescence in situ hybridization targeting TP53, although only one carcinoma harbored a point mutation in this gene. Strikingly, homozygous deletions at 10q23.31, encompassing the PTEN locus, were seen in 58% of the cases with 10q loss.

CONCLUSION

We conclude that aCGH can significantly improve the detection of genomic aberrations in cancer cells as compared to previously established whole-genome methodologies, although contamination with normal cells may influence the sensitivity and specificity of some scoring approaches. Our work delineated recurrent copy number changes and revealed novel amplified loci and frequent homozygous deletions in primary prostate carcinomas, which may guide future work aimed at identifying the relevant target genes. In particular, biallelic loss seems to be a frequent mechanism of inactivation of the PTEN gene in prostate carcinogenesis.

High-resolution genomic profiling of chromosomal aberrations using Infinium whole-genome genotyping

Array-CGH is a powerful tool for the detection of chromosomal aberrations. The introduction of high-density SNP genotyping technology to genomic profiling, termed SNP-CGH, represents a further advance, since simultaneous measurement of both signal intensity variations and changes in allelic composition makes it possible to detect both copy number changes and copy-neutral loss-of-heterozygosity (LOH) events. We demonstrate the utility of SNP-CGH with two Infinium whole-genome genotyping BeadChips, assaying 109,000 and 317,000 SNP loci, to detect chromosomal aberrations in samples bearing constitutional aberrations as well tumor samples at sub-100 kb effective resolution. Detected aberrations include homozygous deletions, hemizygous deletions, copy-neutral LOH, duplications, and amplifications. The statistical ability to detect common aberrations was modeled by analysis of an X chromosome titration model system, and sensitivity was modeled by titration of gDNA from a tumor cell with that of its paired normal cell line. Analysis was facilitated by using a genome browser that plots log ratios of normalized intensities and allelic ratios along the chromosomes. We developed two modes of SNP-CGH analysis, a single sample and a paired sample mode. The single sample mode computes log intensity ratios and allelic ratios by referencing to canonical genotype clusters generated from approximately 120 reference samples, whereas the paired sample mode uses a paired normal reference sample from the same individual. Finally, the two analysis modes are compared and contrasted for their utility in analyzing different types of input gDNA: low input amounts, fragmented gDNA, and Phi29 whole-genome pre-amplified DNA.

STAC: A method for testing the significance of DNA copy number aberrations across multiple array-CGH experiments

Regions of gain and loss of genomic DNA occur in many cancers and can drive the genesis and progression of disease. These copy number aberrations (CNAs) can be detected at high resolution by using microarray-based techniques. However, robust statistical approaches are needed to identify nonrandom gains and losses across multiple experiments/samples. We have developed a method called Significance Testing for Aberrant Copy number (STAC) to address this need. STAC utilizes two complementary statistics in combination with a novel search strategy. The significance of both statistics is assessed, and P-values are assigned to each location on the genome by using a multiple testing corrected permutation approach. We validate our method by using two published cancer data sets. STAC identifies genomic alterations known to be of clinical and biological significance and provides statistical support for 85% of previously reported regions. Moreover, STAC identifies numerous additional regions of significant gain/loss in these data that warrant further investigation. The P-values provided by STAC can be used to prioritize regions for follow-up study in an unbiased fashion. We conclude that STAC is a powerful tool for identifying nonrandom genomic amplifications and deletions across multiple experiments. A Java version of STAC is freely available for download at http://cbil.upenn.edu/STAC.

Robust estimation of the false discovery rate

MOTIVATION

Presently available methods that use p-values to estimate or control the false discovery rate (FDR) implicitly assume that p-values are continuously distributed and based on two-sided tests. Therefore, it is difficult to reliably estimate the FDR when p-values are discrete or based on one-sided tests.

RESULTS

A simple and robust method to estimate the FDR is proposed. The proposed method does not rely on implicit assumptions that tests are two-sided or yield continuously distributed p-values. The proposed method is proven to be conservative and have desirable large-sample properties. In addition, the proposed method was among the best performers across a series of 'real data simulations' comparing the performance of five currently available methods.

AVAILABILITY

Libraries of S-plus and R routines to implement the method are freely available from www.stjuderesearch.org/depts/biostats.

Single-cell chromosomal imbalances detection by array CGH

Genomic imbalances are a major cause of constitutional and acquired disorders. Therefore, aneuploidy screening has become the cornerstone of preimplantation, prenatal and postnatal genetic diagnosis, as well as a routine aspect of the diagnostic workup of many acquired disorders. Recently, array comparative genomic hybridization (array CGH) has been introduced as a rapid and high-resolution method for the detection of both benign and disease-causing genomic copy-number variations. Until now, array CGH has been performed using a significant quantity of DNA derived from a pool of cells. Here, we present an array CGH method that accurately detects chromosomal imbalances from a single lymphoblast, fibroblast and blastomere within a single day. Trisomy 13, 18, 21 and monosomy X, as well as normal ploidy levels of all other chromosomes, were accurately determined from single fibroblasts. Moreover, we showed that a segmental deletion as small as 34 Mb could be detected. Finally, we demonstrated the possibility to detect aneuploidies in single blastomeres derived from preimplantation embryos. This technique offers new possibilities for genetic analysis of single cells in general and opens the route towards aneuploidy screening and detection of unbalanced translocations in preimplantation embryos in particular.

Genomic alterations identified by array comparative genomic hybridization as prognostic markers in tamoxifen-treated estrogen receptor-positive breast cancer

BACKGROUND

A considerable proportion of estrogen receptor (ER)-positive breast cancer recurs despite tamoxifen treatment, which is a serious problem commonly encountered in clinical practice. We tried to find novel prognostic markers in this subtype of breast cancer.

METHODS

We performed array comparative genomic hybridization (CGH) with 1,440 human bacterial artificial chromosome (BAC) clones to assess copy number changes in 28 fresh-frozen ER-positive breast cancer tissues. All of the patients included had received at least 1 year of tamoxifen treatment. Nine patients had distant recurrence within 5 years (Recurrence group) of diagnosis and 19 patients were alive without disease at least 5 years after diagnosis (Non-recurrence group).

RESULTS

Potential prognostic variables were comparable between the two groups. In an unsupervised clustering analysis, samples from each group were well separated. The most common regions of gain in all samples were 1q32.1, 17q23.3, 8q24.11, 17q12-q21.1, and 8p11.21, and the most common regions of loss were 6q14.1-q16.3, 11q21-q24.3, and 13q13.2-q14.3, as called by CGH-Explorer software. The average frequency of copy number changes was similar between the two groups. The most significant chromosomal alterations found more often in the Recurrence group using two different statistical methods were loss of 11p15.5-p15.4, 1p36.33, 11q13.1, and 11p11.2 (adjusted p values < 0.001). In subgroup analysis according to lymph node status, loss of 11p15 and 1p36 were found more often in Recurrence group with borderline significance within the lymph node positive patients (adjusted p = 0.052).

CONCLUSION

Our array CGH analysis with BAC clones could detect various genomic alterations in ER-positive breast cancers, and Recurrence group samples showed a significantly different pattern of DNA copy number changes than did Non-recurrence group samples.

BioHMM: a heterogeneous hidden Markov model for segmenting array CGH data

SUMMARY

We have developed a new method (BioHMM) for segmenting array comparative genomic hybridization data into states with the same underlying copy number. By utilizing a heterogeneous hidden Markov model, BioHMM incorporates relevant biological factors (e.g. the distance between adjacent clones) in the segmentation process.

CARAT: a novel method for allelic detection of DNA copy number changes using high density oligonucleotide arrays

Background

DNA copy number alterations are one of the main characteristics of the cancer cell karyotype and can contribute to the complex phenotype of these cells. These alterations can lead to gains in cellular oncogenes as well as losses in tumor suppressor genes and can span small intervals as well as involve entire chromosomes. The ability to accurately detect these changes is central to understanding how they impact the biology of the cell.

Results

We describe a novel algorithm called CARAT (Copy Number Analysis with Regression And Tree) that uses probe intensity information to infer copy number in an allele-specific manner from high density DNA oligonuceotide arrays designed to genotype over 100, 000 SNPs. Total and allele-specific copy number estimations using CARAT are independently evaluated for a subset of SNPs using quantitative PCR and allelic TaqMan reactions with several human breast cancer cell lines. The sensitivity and specificity of the algorithm are characterized using DNA samples containing differing numbers of X chromosomes as well as a test set of normal individuals. Results from the algorithm show a high degree of agreement with results from independent verification methods.

Conclusion

Overall, CARAT automatically detects regions with copy number variations and assigns a significance score to each alteration as well as generating allele-specific output. When coupled with SNP genotype calls from the same array, CARAT provides additional detail into the structure of genome wide alterations that can contribute to allelic imbalance.

Human and mouse oligonucleotide-based array CGH

Array-based comparative genomic hybridization is a high resolution method for measuring chromosomal copy number changes. Here we present a validated protocol using in-house spotted oligonucleotide libraries for array comparative genomic hybridization (CGH). This oligo array CGH platform yields reproducible results and is capable of detecting single copy gains, multi-copy amplifications as well as homozygous and heterozygous deletions as small as 100 kb with high resolution. A human oligonucleotide library was printed on amine binding slides. Arrays were hybridized using a hybstation and analysed using BlueFuse feature extraction software, with >95% of spots passing quality control. The protocol allows as little as 300 ng of input DNA and a 90% reduction of Cot-1 DNA without compromising quality. High quality results have also been obtained with DNA from archival tissue. Finally, in addition to human oligo arrays, we have applied the protocol successfully to mouse oligo arrays. We believe that this oligo-based platform using ‘off-the-shelf’ oligo libraries provides an easy accessible alternative to BAC arrays for CGH, which is cost-effective, available at high resolution and easily implemented for any sequenced organism without compromising the quality of the results.

A comparison study: applying segmentation to array CGH data for downstream analyses

MOTIVATION

Array comparative genomic hybridization (CGH) allows detection and mapping of copy number of DNA segments. A challenge is to make inferences about the copy number structure of the genome. Several statistical methods have been proposed to determine genomic segments with different copy number levels. However, to date, no comprehensive comparison of various characteristics of these methods exists. Moreover, the segmentation results have not been utilized in downstream analyses.

RESULTS

We describe a comparison of three popular and publicly available methods for the analysis of array CGH data and we demonstrate how segmentation results may be utilized in the downstream analyses such as testing and classification, yielding higher power and prediction accuracy. Since the methods operate on individual chromosomes, we also propose a novel procedure for merging segments across the genome, which results in an interpretable set of copy number levels, and thus facilitate identification of copy number alterations in each genome.

AVAILABILITY

http://www.bioconductor.org