Metabolomics,Unknown,Transcriptomics,Genomics,Proteomics

Dataset Information

A comparison of DNA copy number profiling platforms using a panel of melanoma cell lines

ABSTRACT: The accurate mapping of recurring DNA copy number aberrations (CNAs), a hallmark feature of the cancer genome, has facilitated the discovery of tumor suppressor genes and oncogenes. Microarray-based assays designed to detect these chromosomal copy number alterations on a genome-wide and high-resolution scale have emerged as a cornerstone technology in the genomic era. The diversity of commercially-available platforms prompted a systematic comparison of five copy number profiling assays for their ability to detect 2-fold copy number gain and loss (4n or 1n, respectively) as well as focal high-amplitude CNAs. Here, using a collection of established human melanoma cell lines, we defined the reproducibility, absolute signals, signal:noise, false-positive and false-negative rates for each of the five assays against ground-truth defined by Spectral Karyotyping (SKY), in addition to comparing the concordance of CNAs detection by two high-resolution Agilent and Affymetrix microarray platforms. Our analyses concluded that the Agilentâs 60mer oligo-microarray with probe design optimized for genomic hybridization offers the highest sensitivity and specificity [area under Receiver Operator Characteristic (ROC) curve >0.99] while Affymetrixâs SNP microarray appears to offer better detection of CNAs in gene-poor region. Availability of these comparison results should guide study design decisions and facilitate further computational development. Keywords: comparative genomic hybridization We generated copy number profiles of a defined set of tumor cell lines on five oligo-microarray-based assays of three platforms (Agilent, Affymetrix and NimbleGen) and determined the reproducibility, signal and noise as well as sensitivity and specificity of each in detecting two-fold signals based on SKY-defined aberrations as ground-truth for comparison. In addition, high-density microarray assays from Agilent and Affymetrix platforms were further compared for definition of CNAs in an independent dataset using published analytical approaches. All arrays were run from the same DNA harvest from respective cell lines. Each array was run according to the protocol of the manufacturer.

ORGANISM(S): Homo sapiens

SUBMITTER: Joel Greshock

PROVIDER: E-GEOD-7822 | biostudies-arrayexpress |

REPOSITORIES: biostudies-arrayexpress

ACCESS DATA

Publications

A comparison of DNA copy number profiling platforms.

Greshock Joel J Feng Bin B Nogueira Cristina C Ivanova Elena E Perna Ilana I Nathanson Katherine K Protopopov Alexei A Weber Barbara L BL Chin Lynda L

Cancer research 20071029 21

The accurate mapping of recurring DNA copy number aberrations (CNAs), a hallmark feature of the cancer genome, has facilitated the discovery of tumor suppressor genes and oncogenes. Microarray-based assays designed to detect these chromosomal copy number alterations on a genome-wide and high-resolution scale have emerged as a cornerstone technology in the genomic era. The diversity of commercially available platforms prompted a systematic comparison of five copy number profiling assays for their ...[more]

PMID: 17968032

Similar Datasets

Project description:A common aim of pharmacogenomic studies that employ genome-wide assays on panels of cancers is the unbiased discovery of genomic alterations that are associated with clinical outcome and drug response. Previous investigations of lapatinib, a selective dual-kinase inhibitor of EGFR and HER2 tyrosine kinases, have demonstrated predictable relationships between the activity of these genes and response. Under the hypothesis that additional genes may play a role in promoting sensitivity, a predictive model for lapatinib sensitivity was constructed from genome-wide DNA copy number data from 24 cancer cell lines. An optimal predictive model, which consists of aberrations at nine distinct genetic loci, includes gains of HER2, EGFR, and loss of CDKN2A. This model, which achieved area under the Receiver Operating Characteristic (ROC) curve of ~0.85 (80% confidence interval: 0.70–0.98; P<0.01), and correctly classified the sensitivity status of 8/10 head and neck cancer cell lines. This study demonstrates that previously described biomarkers of lapatinib sensitivity, including copy number gains of EGFR and HER2, can be discovered as powerful predictors of response using novel genomic assays in an unbiased manner. Further, these results demonstrate the utility of DNA copy number profiles in pharmacogenomic studies. Keywords: Comparative Genomic Hybridization Basal DNA copy number profiles were derived from Affymetrix 'SNP chips' for a panel of 34 tumor derived cell lines. Each of these cell lines were tested for their degree of sensitivity to lapatinib. Specific DNA alterations associating with lapatinib sensitivity were identified using a training set consisting of 24 of these lines. A predictive model was constructed from these loci. The remaining 10 lines were used to test the model performance. A panel of chips (n = 12) assayed with germline DNA served as a reference in order to calculate the test/reference ratios reported in the Sample table VALUE columns. The 'reference' Samples for the 500K and 100K assays are as follows: 500K: GSM188024, GSM188025, GSM188026, GSM188027, GSM188028, GSM188029, GSM188030, GSM188031, GSM188032, GSM188033, GSM188034, GSM188035, GSM188048, GSM188049, GSM188050, GSM188051, GSM188052, GSM188053, GSM188054, GSM188055, GSM188056, GSM188057, GSM188058, GSM188059 100K: GSM187950, GSM187951, GSM187952, GSM187953, GSM187954, GSM187955, GSM187956, GSM187957, GSM187958, GSM187959, GSM187960, GSM187961, GSM187949, GSM187948, GSM187946, GSM187947, GSM187945, GSM187944, GSM187943, GSM187942, GSM187941, GSM187940, GSM187939, GSM187938

Project description:Primitive neuro-ectodermal tumours (PNET) of the supratentorial region are rare, highly malignant embryonal brain tumours affecting young children. Although supratentorial PNET (sPNET) are histologically similar to infratentorial PNET/medulloblastoma, sPNET have more aggressive clinical phenotypes, which suggest sPNET represents distinct biological entities. In contrast to considerable progress in understanding the signalling pathways involved in medulloblastoma, little is known about sPNET pathogenesis. Prior low resolution CGH (comparative genomic hybridization) studies indicate sPNET have frequent genomic imbalances and copy number aberrations (CNAs). To define genes involved in sPNET pathogenesis, we utilized the Affymetrix 250K Nsp SNP (single nucleotide polymorphism) analysis to identify genes targeted by recurrent CNAs in primary human sPNET samples. Copy number analysis was conducted on 39 primary PNET samples. Select target genes were validated by genomic and/or RT-PCR. Our analysis revealed frequent CNA across the sPNET genome, encompassing large and focal chromosome segments, and corroborated previous reports that isochromosome 17q, an abnormality found in ~ 30% of medulloblastoma, is rare in sPNET. Keywords: single nucleotide polymorphism array, disease state analysis A total of 56 primary sPNET samples were collected for this study from The Hospital for Sick Children, John Hopkins University, St. Jude Research Hospital, University of Cambridge, Children’s Hospital Boston, Virginia Commonwealth University, Instituto nazionale per lo studio e la cura dei tumori, and Texas Children’s Hospital, with local Institutional Research Ethics Board approval. Pineoblastoma and rhaboid tumours were specifically excluded, leaving 52 samples, 39 of which had good quality DNA available. These samples, in addition to 28 diploid reference samples (a gift from Dr. S. Scherer, University of Toronto), were analyzed on the Affymetrix GeneChip Mapping 250K Nsp SNP array. Due to sample availability at the time the arrays were run, the diploid reference samples analyzed on the Sty arrays are different from reference samples analyzed on the Nsp arrays.

Project description:Several copy number altered regions (CNA) have been identified in the genome of cervical cancer, especially amplifications of 3q and 5p. However, the contribution of those alterations to cervical carcinogenesis is still a matter of debate, since genome-wide, there is a lack of correlation between CNAs and gene expression. In this study, we investigated whether the CNAs in cell lines (CaLo, CasKi, HeLa, SiHa), at a gene-by-gene level, are related to changes in gene expression. On average 19.2% of the whole genome of cell lines had CNA. However, only 2.4% consisted of minimal recurrent regions (MRR), common to all cell lines. Whereas 3q had just some sparse common gains (13%), 5p was entirely duplicated recurrently. Genome-wide, only 11% of genes located in CNAs changed gene expression. In contrast, the rate increased over 3 fold times in MRRs. Chr 5p was confirmed entirely amplified by FISH. In spite of this, at most 32.9% of the explored genes in 5p (n=202) were de-regulated. In 3q, the rate was just 11.8%. Even in 3q26, which had five MRRs and 38.7% of SNPs was gained recurrently, the rate rose slightly to 13.6% (10 out of 73). Interestingly, up to 16% of de-regulated genes in 5p and 80% in 3q26 were down-regulated, suggesting additional factors are involved in gene repression. The de-regulated genes in 3q and 5p were found in clusters, suggesting local chromatin factors may also influence gene expression. In regions amplified discontinuously, the rate of down-regulated genes rose steadily as the number of amplified SNPs increased (p<0.01, Spearman's correlation). This suggests partial gene amplification as a mechanism of silencing gene expression. Additional genes were identified up- or down-regulated in 5p and 3q, which could be involved in cervical carcinogenesis, especially implicated in apoptosis. Those include CLPTM1L, AHRR, PDCD6 and DAP in 5p and TNFSF10 and ECT2 in 3q. Overall, the gene expression and copy number profiles suggest other factors, like epigenetic or chromatin domains, may influence gene expression within the entirely amplified genome segments. Further studies are needed to elucidate how these mechanisms regulate gene expression. Affymetrix SNP arrays were performed according to the manufacturer's directions on DNA extracted from cell lines or peripheral blood samples. Copy number analysis of Affymetrix 100K SNP arrays was performed for 4 cervical cancer cell lines. DNA from 38 healty women were used as references for copy number inference.

Project description:Introduction: A major challenge in the interpretation of genomic profiling data generated from breast cancer samples is the identification of driver genes as distinct from bystander genes which do not impact tumorigenesis. One way to assess the relative importance of alterations in the transcriptome profile is to combine complementary analyses that assess changes in the copy number alterations (CNAs). This integrated analysis permits the identification of genes with altered expression that map within specific chromosomal regions that demonstrate copy number alterations, providing a mechanistic approach to identify the 'driver genes’. Methods: We have performed whole genome analysis of CNAs using the Affymetrix 250K Mapping array on 22 infiltrating ductal carcinoma samples (IDCs). Analysis of transcript expression alterations was performed using the Affymetrix U133 Plus2.0 array on 16 IDC samples. Twelve IDC samples were analyzed using both platforms and the data integrated. We also incorporated data from LOH analysis to identify genes showing loss of expression in LOH regions. Results: Copy number analysis results demarcated smaller boundaries for many previously reported CNAs, and in some cases, the CNAs were defined as more than a single contiguous event. Additionally, we were able to assign driver genes to these commonly reported regions using a rigorous methodology. For example, RAB25 showed a large increased expression in the tumors and mapped to the commonly reported amplification at 1q22. We also identified 5 genes in the 8q24 amplicon and TSEN4 in the 17q25 amplified region. LOH analysis confirmed some previously reported regions, and integration with copy number data determined that the detected LOH were copy neutral events. Finally, we have identified several RXR pathways that demonstrated down-regulation in IDC whose members may represent further targets of therapeutic intervention. Conclusion: We have demonstrated the utility of the application of integrated analysis using high-resolution CGH and whole genome transcript analysis for detecting driver genes in IDC. The high resolution platform allowed a refined demarcation of CNAs, and gene expression profiling provided a mechanism to detect genes directly impacted by the CNA. This is the first report of LOH in IDC using a high resolution platform. 22 IDC samples were analyzed for CNA and an overlapping 16 samples were analyzed for gene expression

Dataset Information

A comparison of DNA copy number profiling platforms using a panel of melanoma cell lines

Publications

A comparison of DNA copy number profiling platforms.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure