Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “unsupervised clustering”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 253 records · Page 14Linked to original sources
PubMed2005

Molecular characterization of salivary gland malignancy using the Smgb-Tag transgenic mouse model.

The molecular mechanisms underlying salivary gland tumorigenesis remain unclear. In order to identify genetic changes that occur during the development of invasive adenocarcinoma from normal salivary gland, we used the Smgb-Tag transgenic mouse model. This transgene induces the progressive development of dysplasia to invasive adenocarcinoma in the submandibular salivary gland. Gene expression patterns from 20 submandibular glands (two normal, nine dysplasia and nine adenocarcinoma samples) were assessed using a mouse 15 K cDNA array. Unsupervised hierarchical clustering was used to group gene expression based on 157 differentially expressed genes distinguishing between dysplasias and adenocarcinomas. Further analysis identified 25 significantly overexpressed and 28 underexpressed cDNA sequences in adenocarcinoma as compared to dysplasia. Differential expression of five genes (Lcn2, Ptn, Cd24a, Mapk6 and Rnps1) was validated by quantitative real-time RT-PCR in a total of 48 mouse salivary gland tissues (seven histologically normal, 13 dysplasias and 28 adenocarcinomas), including the 20 samples analyzed by cDNA arrays. Immunohistochemical analysis was used to validate the expression of Ptn and Cd24a at the protein level in a subset of 16 mouse salivary glands (four normal, five dysplasia and seven adenocarcinoma samples), as well as in 23 human submandibular gland tumors (16 pleomorphic adenomas, three adenoid cystic carcinomas, one acinic cell carcinoma, one adenocarcinoma NOS, one myoepithelial and one mucoepidermoid carcinoma). We thus demonstrated that the Smgb-Tag transgenic mouse model is a useful tool for the identification of genes that are deregulated in salivary gland adenocarcinomas. Our data suggest that Ptn and Cd24a may be genetic markers associated with salivary gland tumorigenesis and/or progression.

Adenocarcinoma↗
PubMed2006

Identification of pathways regulating cell size and cell-cycle progression by RNAi.

Many high-throughput loss-of-function analyses of the eukaryotic cell cycle have relied on the unicellular yeast species Saccharomyces cerevisiae and Schizosaccharomyces pombe. In multicellular organisms, however, additional control mechanisms regulate the cell cycle to specify the size of the organism and its constituent organs. To identify such genes, here we analysed the effect of the loss of function of 70% of Drosophila genes (including 90% of genes conserved in human) on cell-cycle progression of S2 cells using flow cytometry. To address redundancy, we also targeted genes involved in protein phosphorylation simultaneously with their homologues. We identify genes that control cell size, cytokinesis, cell death and/or apoptosis, and the G1 and G2/M phases of the cell cycle. Classification of the genes into pathways by unsupervised hierarchical clustering on the basis of these phenotypes shows that, in addition to classical regulatory mechanisms such as Myc/Max, Cyclin/Cdk and E2F, cell-cycle progression in S2 cells is controlled by vesicular and nuclear transport proteins, COP9 signalosome activity and four extracellular-signal-regulated pathways (Wnt, p38betaMAPK, FRAP/TOR and JAK/STAT). In addition, by simultaneously analysing several phenotypes, we identify a translational regulator, eIF-3p66, that specifically affects the Cyclin/Cdk pathway activity.

Active Transport, Cell Nucleus↗
PubMed2005

Array comparative genomic hybridisation (aCGH) analysis of premenopausal breast cancers from a nuclear fallout area and matched cases from Western New York.

High-resolution array comparative genomic hybridisation (aCGH) analysis of DNA copy number aberrations (CNAs) was performed on breast carcinomas in premenopausal women from Western New York (WNY) and from Gomel, Belarus, an area exposed to fallout from the 1986 Chernobyl nuclear accident. Genomic DNA was isolated from 47 frozen tumour specimens from 42 patients and hybridised to arrays spotted with more than 3000 BAC clones. In all, 20 samples were from WNY and 27 were from Belarus. In total, 34 samples were primary tumours and 13 were lymph node metastases, including five matched pairs from Gomel. The average number of total CNAs per sample was 76 (range 35-134). We identified 152 CNAs (92 gains and 60 losses) occurring in more than 10% of the samples. The most common amplifications included gains at 8q13.2 (49%), at 1p21.1 (36%), and at 8q24.21 (36%). The most common deletions were at 1p36.22 (26%), at 17p13.2 (26%), and at 8p23.3 (23%). Belarussian tumours had more amplifications and fewer deletions than WNY breast cancers. HER2/neu negativity and younger age were also associated with a higher number of gains and fewer losses. In the five paired samples, we observed more discordant than concordant DNA changes. Unsupervised hierarchical cluster analysis revealed two distinct groups of tumours: one comprised predominantly of Belarussian carcinomas and the other largely consisting of WNY cases. In total, 50 CNAs occurred significantly more commonly in one cohort vs the other, and these included some candidate signature amplifications in the breast cancers in women exposed to significant radiation. In conclusion, our high-density aCGH study has revealed a large number of genetic aberrations in individual premenopausal breast cancer specimens, some of which had not been reported before. We identified a distinct CNA profile for carcinomas from a nuclear fallout area, suggesting a possible molecular fingerprint of radiation-associated breast cancer.

Adult↗
PubMed2004

Gene expression-based, individualized outcome prediction for surgically treated lung cancer patients.

Individualized outcome prediction classifiers were successfully constructed through expression profiling of a total of 8644 genes in 50 non-small-cell lung cancer (NSCLC) cases, which had been consecutively operated on within a defined short period of time and followed up for more than 5 years. The resultant classifier of NSCLCs yielded 82% accuracy for forecasting survival or death 5 years after surgery of a given patient. In addition, since two major histologic classes may differ in terms of outcome-related expression signatures, histologic-type-specific outcome classifiers were also constructed. The resultant highly predictive classifiers, designed specifically for nonsquamous cell carcinomas, showed a prediction accuracy of more than 90% independent of disease stage. In addition to the presence of heterogeneities in adenocarcinomas, our unsupervised hierarchical clustering analysis revealed for the first time the existence of clinicopathologically relevant subclasses of squamous cell carcinomas with marked differences in their invasive growth and prognosis. This finding clearly suggests that NSCLCs comprise distinct subclasses with considerable heterogeneities even within one histologic type. Overall, these findings should advance not only our understanding of the biology of lung cancer but also our ability to individualize postoperative therapies based on the predicted outcome.

Adenocarcinoma↗
PubMed2004

Gastrointestinal stromal tumors (GISTs) with KIT and PDGFRA mutations have distinct gene expression profiles.

Most GISTs require oncogenic activation of the KIT or PDGFRA receptor tyrosine kinase proteins, and the genomic mechanisms of oncogene activation are heterogeneous. Notably, the kinase mutation type correlates with both tumor biology and imatinib response. For example, GISTs with KIT exon 11 mutations are typically gastric and have excellent imatinib response, whereas those with KIT exon 9 mutations generally arise in the small bowel and are less responsive to imatinib. To identify genes that might contribute to these biological differences, we carried out gene expression profiling of 26 GISTs with known KIT and PDGFRA mutational status. Expression differences were then evaluated further by RNA in situ hybridization, immunohistochemistry, and immunoblotting. Unsupervised hierarchical clustering grouped tumors with similar mutations together, but the distinction between the different groups was not absolute. Differentially expressed genes included ezrin, p70S6K, and PKCs, which are known to have key roles in KIT or PDGFRA signaling, and which might therefore contribute to the distinctive clinicopathological features in GISTs with different mutation types. These gene products could serve as highly selective therapeutic targets in GISTs containing the KIT or PDGFRA mutational types with which they are associated.

Adolescent↗
PubMed2006

Serrated carcinomas form a subclass of colorectal cancer with distinct molecular basis.

Serrated colorectal carcinomas (CRCs) are morphologically different from conventional CRCs and have been proposed to follow a distinct pathway of CRC formation. Despite studies of single molecular events in this tumor type, the diagnosis of serrated CRC relies on morphology and the putative unique biological character of these tumors has not been established. Here we show that the gene expression profiling of 37 CRCs separated serrated and conventional CRCs into two distinct branches in unsupervised hierarchical clustering (P-value 7.8 x 10(-7)), and revealed 201 differentially expressed genes representing potential biomarkers for serrated CRC. Immunohistochemistry was utilized to verify the key findings in the 37 CRCs examined by expression profiling, and a separate validation set of 37 serrated and 86 conventional CRCs was examined to evaluate the candidate biomarkers in an extended sample material. Ephrin receptor B2, hypoxia-inducible factor 1-alpha and patched appeared as proteins important for genesis of serrated CRC. This study establishes serrated CRCs as a biologically distinct subclass of CRC and represents a step forward in the molecular classification of these cancers. The study also provides a platform to understand the molecular basis of serrated CRC and in long term may contribute to the development of specific treatment options for this tumor type.

Adenocarcinoma, Mucinous↗
PubMed2006

Using array-comparative genomic hybridization to define molecular portraits of primary breast cancers.

We analysed 148 primary breast cancers using BAC-arrays containing 287 clones representing cancer-related gene/loci to obtain genomic molecular portraits. Gains were detected in 136 tumors (91.9%) and losses in 123 tumors (83.1%). Eight tumors (5.4%) did not have any genomic aberrations in the 281 clones analysed. Common (more than 15% of the samples) gains were observed at 8q11-qtel, 1q21-qtel, 17q11-q12 and 11q13, whereas common losses were observed at 16q12-qtel, 11ptel-p15.5, 1p36-ptel, 17p11.2-p12 and 8ptel-p22. Patients with tumors registering either less than 5% (median value) or less than 11% (third quartile) total copy number changes had a better overall survival (log-rank test: P=0.0417 and P=0.0375, respectively). Unsupervised hierarchical clustering based on copy number changes identified four clusters. Women with tumors from the cluster with amplification of three regions containing known breast oncogenes (11q13, 17q12 and 20q13) had a worse prognosis. The good prognosis group (Nottingham Prognostic Index (NPI) <or=3.4) tumors had frequent loss of 16q24-qtel. Genes significantly associated with estrogen receptor (ER), Grade and NPI were used to build k-nearest neighbor (KNN) classifiers that predicted ER, Grade and NPI status in the test set with an average misclassification rate of 24.7, 25.7 and 35.7%, respectively. These data raise the prospect of generating a molecular taxonomy of breast cancer based on copy number profiling using tumor DNA, which may be more generally applicable than expression microarray analysis.

Breast Neoplasms↗
PubMed2006

Identification of molecular markers and signaling pathway in endometrial cancer in Hong Kong Chinese women by genome-wide gene expression profiling.

Endometrial cancer is the third most common gynecologic malignancy and the ninth most common malignancy for females overall in Hong Kong. Approximately 80% or more of these cancers are endometrioid endometrial adenocarcinomas. The aim of this study was to reveal genes contributing to the development of endometrioid endometrial cancer, which may impact diagnosis, prognosis and treatment of the disease. Whole-genome gene expression analysis was completed for a set of 55 microdissected sporadic endometrioid endometrial adenocarcinomas and 29 microdissected normal endometrium specimens using the Affymetrix Human U133 Plus 2.0 oligonucleotide microarray. Selected genes of interest were validated by quantitative real-time-polymerase chain reaction (qRT-PCR). Pathway analysis was performed to reveal gene interactions involved in endometrial tumorigenesis. Unsupervised hierarchical clustering displayed a distinct separation between the endometrioid adenocarcinomas and normal endometrium samples. Supervised analysis identified 117 highly differentially regulated genes (>or=4.0-fold change), which distinguished the endometrial cancer specimens from normal endometrium. Twelve novel genes including DKK4, ZIC1, KIF1A, SAA2, LOC16378, ALPP2, CCL20, CXCL5, BST2, OLFM1, KLRC1 and MBC45780 were deregulated in the endometrial cancer, and further validated in an independent set of 56 cancer and 29 normal samples using qRT-PCR. In addition, 10 genes were differentially regulated in late-stage cancer, as compared to early-stage disease, and may be involved in tumor progression. Pathway analysis of the expression data from this tumor revealed an interconnected network consisting of 21 aberrantly regulated genes involved in angiogenesis, cell proliferation and chromosomal instability. The results of this study highlight the molecular features of endometrioid endometrial cancer and provide insight into the events underlying the development and progression of endometrioid endometrial cancer.

Endometrial Neoplasms↗
PubMed2002

Screening of gene expression profiles in gastric epithelial cells induced by Helicobacter pylori using microarray analysis.

BACKGROUND: H. pylori infection is a major risk factor in gastric cancer development. The availability of cDNA microarrays creates the unprecedented opportunity to examine simultaneously dynamic changes of multiple pathways affected by H. pylori infection. AIM: In this study we examined broad patterns of gene expression induced by H. pylori in the gastric cancer cell line 1739-CRL AGS cells in culture using the U95A microarray. METHODS: H. pylori were cocultured with AGS cells for 4, 12, 24 and 48 h. Total RNA was extracted and after labelling was used for detection of genes represented in the human U95A microarray set. Data analyses were performed using GeneChip and CLUSFAVOR software. RESULTS: Nearly 6000 genes present in the array were expressed by AGS cells. We report approximately 200 genes that showed the most marked changes. Our studies confirm the up-regulation of c-jun, jun-B, c-fos and cyclin D1 by H. pylori. We report for the first time the induction of the serine threonine kinase pim-1 and ATF3 by H. pylori infection of AGS cells. CONCLUSIONS: In this microarray analysis of gene expression induced by H. pylori in gastric epithelial cells, we identified a large number of unsuspected genes affected by H. pylori. Further, we show that unsupervised hierarchical cluster analysis can provide useful insight into the possible contribution of genes in specific pathways, based on their profile of expression.

Activating Transcription Factor 3↗
PubMed2005

Gene-expression profiling predicts recurrence in Dukes' C colorectal cancer.

BACKGROUND & AIMS: Although approximately 50% of Dukes' C colorectal cancer patients are surgically cured, it is currently not possible to distinguish these patients from those at high risk of recurrence. The recent advent of routine adjuvant chemotherapy for these patients has greatly complicated the identification of new markers predicting the response to surgery, which is now reliant on archived materials. Microarray analysis allows fine tumor classification but cannot be used with paraffin-embedded archival samples. METHODS: We used microarray analysis of a unique set of fresh-frozen tumor samples from Dukes' C patients who had surgery as the only form of treatment to identify molecular signatures that characterize tumors from patients with good and bad prognosis. RESULTS: Unsupervised hierarchical clustering and a K-nearest neighbors-based classifier identified groups of patients with significantly different survival (P = .019 and P = .0001). Expression profiling outperformed previously reported genetic markers of prognosis such as TP53 and K-RAS mutational status and allelic imbalance in chromosome 18q, which were of limited prognostic power in this study. Functional categories significantly enriched in gene-expression differences included protein transport and folding. The prognostic potential of the RAS homologue RHOA, one of the most differentially expressed genes, was further investigated using immunohistochemistry and a tissue microarray containing 137 independent Dukes' C tumor samples. Reduced RHOA expression was associated with significantly shorter survival (P = .01). CONCLUSIONS: This study shows that gene-expression profiling of surgical tumor samples can predict recurrence in Dukes' C patients. Therefore, this approach could be used to guide decisions concerning the clinical management of these patients.

Aged↗
PubMed2004

Gene expression profiling identifies clinically relevant subtypes of prostate cancer.

Prostate cancer, a leading cause of cancer death, displays a broad range of clinical behavior from relatively indolent to aggressive metastatic disease. To explore potential molecular variation underlying this clinical heterogeneity, we profiled gene expression in 62 primary prostate tumors, as well as 41 normal prostate specimens and nine lymph node metastases, using cDNA microarrays containing approximately 26,000 genes. Unsupervised hierarchical clustering readily distinguished tumors from normal samples, and further identified three subclasses of prostate tumors based on distinct patterns of gene expression. High-grade and advanced stage tumors, as well as tumors associated with recurrence, were disproportionately represented among two of the three subtypes, one of which also included most lymph node metastases. To further characterize the clinical relevance of tumor subtypes, we evaluated as surrogate markers two genes differentially expressed among tumor subgroups by using immunohistochemistry on tissue microarrays representing an independent set of 225 prostate tumors. Positive staining for MUC1, a gene highly expressed in the subgroups with "aggressive" clinicopathological features, was associated with an elevated risk of recurrence (P = 0.003), whereas strong staining for AZGP1, a gene highly expressed in the other subgroup, was associated with a decreased risk of recurrence (P = 0.0008). In multivariate analysis, MUC1 and AZGP1 staining were strong predictors of tumor recurrence independent of tumor grade, stage, and preoperative prostate-specific antigen levels. Our results suggest that prostate tumors can be usefully classified according to their gene expression patterns, and these tumor subtypes may provide a basis for improved prognostication and treatment stratification.

Biomarkers, Tumor↗
PubMed2005

Robustness, scalability, and integration of a wound-response gene expression signature in predicting breast cancer survival.

Based on the hypothesis that features of the molecular program of normal wound healing might play an important role in cancer metastasis, we previously identified consistent features in the transcriptional response of normal fibroblasts to serum, and used this "wound-response signature" to reveal links between wound healing and cancer progression in a variety of common epithelial tumors. Here, in a consecutive series of 295 early breast cancer patients, we show that both overall survival and distant metastasis-free survival are markedly diminished in patients whose tumors expressed this wound-response signature compared to tumors that did not express this signature. A gene expression centroid of the wound-response signature provides a basis for prospectively assigning a prognostic score that can be scaled to suit different clinical purposes. The wound-response signature improves risk stratification independently of known clinico-pathologic risk factors and previously established prognostic signatures based on unsupervised hierarchical clustering ("molecular subtypes") or supervised predictors of metastasis ("70-gene prognosis signature").

Breast Neoplasms↗
PubMed2005

The gene expression signatures of melanoma progression.

Because of the paucity of available tissue, little information has previously been available regarding the gene expression profiles of primary melanomas. To understand the molecular basis of melanoma progression, we compared the gene expression profiles of a series of nevi, primary melanomas, and melanoma metastases. We found that metastatic melanomas exhibit two dichotomous patterns of gene expression, which unexpectedly reflect gene expression differences already apparent in comparing laser-capture microdissected radial and vertical phases of a large primary melanoma. Unsupervised hierarchical clustering accurately separated nevi and primary melanomas. Multiclass significance analysis of microarrays comparing normal skin, nevi, primary melanomas, and the two types of metastatic melanoma identified 2,602 transcripts that significantly correlated with sample class. These results suggest that melanoma pathogenesis can be understood as a series of distinct molecular events. The gene expression signatures identified here provide the basis for developing new diagnostics and targeting therapies for patients with malignant melanoma.

Biopsy↗
PubMed2003

Gene expression profiles of primary breast tumors maintained in distant metastases.

It has been debated for decades how cancer cells acquire metastatic capability. It is unclear whether metastases are derived from distinct subpopulations of tumor cells within the primary site with higher metastatic potential, or whether they originate from a random fraction of tumor cells. Here we show, by gene expression profiling, that human primary breast tumors are strikingly similar to the distant metastases of the same patient. Unsupervised hierarchical clustering, multidimensional scaling, and permutation testing, as well as the comparison of significantly expressed genes within a pair, reveal their genetic similarity. Our findings suggest that metastatic capability in breast cancer is an inherent feature and is not based on clonal selection.

Breast Neoplasms↗
PubMed2005

An image-analysis system based on support vector machines for automatic grade diagnosis of brain-tumour astrocytomas in clinical routine.

An image-analysis system based on the concept of Support Vector Machines (SVM) was developed to assist in grade diagnosis of brain tumour astrocytomas in clinical routine. One hundred and forty biopsies of astrocytomas were characterized according to the WHO system as grade II, III and IV. Images from biopsies were digitized, and cell nuclei regions were automatically detected by encoding texture variations in a set of wavelet, autocorrelation and parzen estimated descriptors and using an unsupervised SVM clustering methodology. Based on morphological and textural nuclear features, a decision-tree classification scheme distinguished between different grades of tumours employing an SVM classifier. The system was validated for clinical material collected from two different hospitals. On average, the SVM clustering algorithm correctly identified and accurately delineated 95% of all nuclei. Low-grade tumours were distinguished from high-grade tumours with an accuracy of 90.2% and grade III from grade IV with an accuracy of 88.3% The system was tested in a new clinical data set, and the classification rates were 87.5 and 83.8%, respectively. Segmentation and classification results are very encouraging, considering that the method was developed based on every-day clinical standards. The proposed methodology might be used in parallel with conventional grading to support the regular diagnostic procedure and reduce subjectivity in astrocytomas grading.

Astrocytoma↗
PubMedDate not supplied

Gene expression by PBMC in primary sclerosing cholangitis: evidence for dysregulation of immune mediated genes.

Primary sclerosing cholangitis (PSC) is a chronic disease of the bile ducts characterized by an inflammatory infiltrate and obliterative fibrosis. The precise role of the immune system in the pathogenesis of PSC remains unknown. We used RNA microarray analysis to identify immune-related genes and pathways that are differentially expressed in PSC. Messenger RNA (mRNA) from peripheral blood mononuclear cells (PBMC) was isolated from both patients with PSC and age and sex matched healthy controls. Samples from 5 PSC patients and 5 controls were analyzed by microarray and based upon rigorous statistical analysis of the data, relevant genes were chosen for confirmation by RT-PCR in 10 PSC patients and 10 controls. Using unsupervised hierarchical clustering, gene expression in PSC was statistically different from our control population. Interestingly, genes within the IL-2 receptor beta, IL-6 and MAP Kinase pathways were found to be differently expressed in patients with PSC compared to controls. Further, individual genes, TNF-alpha induced protein 6 (TNFaip6) and membrane-spanning 4-domains, subfamily A (ms4a) were found to be upregulated in PSC while similar to Mothers against decapentaplegic homolog 5 (SMAD 5) was downregulated. In conclusion, several immune-related pathways and genes were differentially expressed in PSC compared to control patients, giving further evidence that this disease is systemic and immune-mediated.

Adult↗
PubMed2004

Development of biomarkers based on diet-dependent metabolic serotypes: practical issues in development of expert system-based classification models in metabolomic studies.

Dietary restriction (DR)-induced changes in the serum metabolome may be biomarkers for physiological status (e.g., relative risk of developing age-related diseases such as cancer). Megavariate analysis (unsupervised hierarchical cluster analysis [HCA]; principal components analysis [PCA]) of serum metabolites reproducibly distinguish DR from ad libitum fed rats. Component-based approaches (i.e., PCA) consistently perform as well as or better than distance-based metrics (i.e., HCA). We therefore tested the following: (A) Do identified subsets of serum metabolites contain sufficient information to construct mathematical models of class membership (i.e., expert systems)? (B) Do component-based metrics out-perform distance-based metrics? Testing was conducted using KNN (k-nearest neighbors, supervised HCA) and SIMCA (soft independent modeling of class analogy, supervised PCA). Models were built with single cohorts, combined cohorts or mixed samples from previously studied cohorts as training sets. Both algorithms over-fit models based on single cohort training sets. KNN models had >85% accuracy within training/test sets, but were unstable (i.e., values of k could not be accurately set in advance). SIMCA models had 100% accuracy within all training sets, 89 % accuracy in test sets, did not appear to over-fit mixed cohort training sets, and did not require post-hoc modeling adjustments. These data indicate that (i) previously defined metabolites are robust enough to construct classification models (expert systems) with SIMCA that can predict unknowns by dietary category; (ii) component-based analyses outperformed distance-based metrics; (iii) use of over-fitting controls is essential; and (iv) subtle inter-cohort variability may be a critical issue for high data density biomarker studies that lack state markers.

Algorithms↗
PubMed2004

Gene expression profiling of chemically induced rat mammary gland cancer.

Exposure to carcinogens through diet, the atmosphere and other means is generally regarded as influencing human cancer risk, but the impact of specific environmental carcinogens on human breast cancer incidence is still unknown. We examined whether distinct chemical carcinogens induce a unique transcriptional profile in mammary gland cancer that is characteristic of the etiologic agent. Rat mammary gland cancers (n = 34) were generated by various carcinogens, including the food-derived heterocyclic amines 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine and 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline, 7,12-dimethylbenz[a]anthracene, N-nitrosomethylurea and 4-aminobiphenyl. The histopathology of the carcinomas was graded using a modified Scarff-Bloom-Richardson scheme and the gene expression profiles in the carcinomas were evaluated on a 10K cDNA microarray. Unsupervised hierarchical clustering analysis revealed two major clusters of carcinomas irrespective of the carcinogenic agent that distinguished two groups with different histopathological parameters (degree of differentiation, nuclear grade, mitotic activity, epithelial cell growth pattern and necrosis). Using class comparison analysis and hierarchical clustering of all carcinomas irrespective of histopathology, gene expression profiles were further shown to be statistically differentially expressed according to the carcinogenic agent. These findings indicate that the transcriptional program in carcinomas is unique to the etiologic agent and can be observed among a diverse set of carcinogens despite variations in carcinoma histopathology. The ability to use microarray analysis to discern an etiology-specific profile among a pathologically heterogeneous group of breast carcinomas may ultimately be valuable in determining the role of environmental chemical carcinogens in human breast cancer risk.

Animals↗