FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2095194002> ?p ?o ?g. }

• W2095194002 endingPage "115" @default.
• W2095194002 startingPage "106" @default.
• W2095194002 abstract "Carcinomas of unknown primary origin constitute 3% to 5% of all newly diagnosed metastatic cancers, with the primary source difficult to classify with current histological methods. Effective cancer treatment depends on early and accurate identification of the tumor; patients with metastases of unknown origin have poor prognosis and short survival. Because miRNA expression is highly tissue specific, the miRNA profile of a metastasis may be used to identify its origin. We therefore evaluated the potential of miRNA profiling to identify the primary tumor of known metastases. Two hundred eight formalin-fixed, paraffin-embedded samples, representing 15 different histologies, were profiled on a locked nucleic acid–enhanced microarray platform, which allows for highly sensitive and specific detection of miRNA. On the basis of these data, we developed and cross-validated a novel classification algorithm, least absolute shrinkage and selection operator, which had an overall accuracy of 85% (CI, 79%–89%). When the classifier was applied on an independent test set of 48 metastases, the primary site was correctly identified in 42 cases (88% accuracy; CI, 75%–94%). Our findings suggest that miRNA expression profiling on paraffin tissue can efficiently predict the primary origin of a tumor and may provide pathologists with a molecular diagnostic tool that can improve their capability to correctly identify the origin of hitherto unidentifiable metastatic tumors and, eventually, enable tailored therapy. Carcinomas of unknown primary origin constitute 3% to 5% of all newly diagnosed metastatic cancers, with the primary source difficult to classify with current histological methods. Effective cancer treatment depends on early and accurate identification of the tumor; patients with metastases of unknown origin have poor prognosis and short survival. Because miRNA expression is highly tissue specific, the miRNA profile of a metastasis may be used to identify its origin. We therefore evaluated the potential of miRNA profiling to identify the primary tumor of known metastases. Two hundred eight formalin-fixed, paraffin-embedded samples, representing 15 different histologies, were profiled on a locked nucleic acid–enhanced microarray platform, which allows for highly sensitive and specific detection of miRNA. On the basis of these data, we developed and cross-validated a novel classification algorithm, least absolute shrinkage and selection operator, which had an overall accuracy of 85% (CI, 79%–89%). When the classifier was applied on an independent test set of 48 metastases, the primary site was correctly identified in 42 cases (88% accuracy; CI, 75%–94%). Our findings suggest that miRNA expression profiling on paraffin tissue can efficiently predict the primary origin of a tumor and may provide pathologists with a molecular diagnostic tool that can improve their capability to correctly identify the origin of hitherto unidentifiable metastatic tumors and, eventually, enable tailored therapy. Although most patients with cancer present with a primary tumor (at its site of origin), 10% to 15% of all cancers are diagnosed as metastases, and one-third of these may have a site of origin, which remains elusive, even after thorough physical and radiological examination, blood tests, and histological evaluation.1Greco F.A. Cancer of unknown primary site.Am Soc Clin Oncol Educ Book. 2013; 2013: 175-181Crossref Scopus (18) Google Scholar Thus, metastatic cancer of unknown primary (CUP) origin accounts for 3% to 6% of all cancer diagnoses and represents the seventh most frequent type of cancer, ranking below cancers of the lung, prostate, breast, cervix, colon, and stomach. Because effective cancer treatment depends on early identification of the primary tumor, patients with CUP origin have a poor prognosis with a median survival of 3 to 6 months and a 1-year survival rate of <25%. In addition, many patients with CUP origin are diagnosed with poorly differentiated adenocarcinomas, which make morphological and immunohistochemical interpretation difficult. Thus, an unrecognized number of patients may be misclassified for tumor origin, and these patients could benefit from improved molecular classification.2Daugaard D. Møller A. Petersen B. Tumors of unknown origin.in: Cavalli F. Kaye S. Hansen H. Armitage J. Piccart-Gebhart M. Textbook of Medical Oncology. Informa, London2009: 313-322Crossref Google Scholar Cancer classification that is based on gene expression profiling by DNA microarrays was reported in 1999 for leukemia by Golub et al3Golub T.R. Slonim D.K. Tamayo P. Huard C. Gaasenbeek M. Mesirov J.P. Coller H. Loh M.L. Downing J.R. Caligiuri M.A. Bloomfield C.D. Lander E.S. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring.Science. 1999; 286: 531-537Crossref PubMed Scopus (9190) Google Scholar and, subsequently, has been extended to include categorization of solid tumors.4Ramaswamy S. Tamayo P. Rifkin R. Mukherjee S. Yeang C.H. Angelo M. Ladd C. Reich M. Latulippe E. Mesirov J.P. Poggio T. Gerald W. Loda M. Lander E.S. Golub T.R. Multiclass cancer diagnosis using tumor gene expression signatures.Proc Natl Acad Sci U S A. 2001; 98: 15149-15154Crossref PubMed Scopus (1646) Google Scholar, 5Buckhaults P. Zhang Z. Chen Y.C. Wang T.L. St Croix B. Saha S. Bardelli A. Morin P.J. Polyak K. Hruban R.H. Velculescu V.E. Shih IeM. Identifying tumor origin using a gene expression-based classification map.Cancer Res. 2003; 63: 4144-4149PubMed Google Scholar, 6Ma X.J. Patel R. Wang X. Salunga R. Murage J. Desai R. Tuggle J.T. Wang W. Chu S. Stecker K. Raja R. Robin H. Moore M. Baunoch D. Sgroi D. Erlander M. Molecular classification of human cancers using a 92-gene real-time quantitative polymerase chain reaction assay.Arch Pathol Lab Med. 2006; 130: 465-473PubMed Google Scholar, 7Kurahashi I. Fujita Y. Arao T. Kurata T. Koh Y. Sakai K. Matsumoto K. Tanioka M. Takeda K. Takiguchi Y. Yamamoto N. Tsuya A. Matsubara N. Mukai H. Minami H. Chayahara N. Yamanaka Y. Miwa K. Takahashi S. Takahashi S. Nakagawa K. Nishio K. A microarray-based gene expression analysis to identify diagnostic biomarkers for unknown primary cancer.PLoS One. 2013; 8: e63249Crossref PubMed Scopus (18) Google Scholar, 8Greco F.A. Lennington W.J. Spigel D.R. Hainsworth J.D. Molecular profiling diagnosis in unknown primary cancer: accuracy and ability to complement standard pathology.J Natl Cancer Inst. 2013; 105: 782-790Crossref PubMed Scopus (76) Google Scholar miRNAs constitute a recently discovered class of tissue-specific, small, noncoding RNAs, which regulate the expression of genes involved in many biological processes, including development, differentiation, apoptosis, and carcinogenesis.9Esquela-Kerscher A. Slack F.J. Oncomirs - microRNAs with a role in cancer.Nat Rev Cancer. 2006; 6: 259-269Crossref PubMed Scopus (6186) Google Scholar, 10Iorio M.V. Croce C.M. microRNA involvement in human cancer.Carcinogenesis. 2012; 33: 1126-1133Crossref PubMed Scopus (487) Google Scholar That miRNAs are promising molecular biomarkers for classification of cancer has previously been suggested by Lu et al,11Lu J. Getz G. Miska E.A. varez-Saavedra E. Lamb J. Peck D. Sweet-Cordero A. Ebert B.L. Mak R.H. Ferrando A.A. Downing J.R. Jacks T. Horvitz H.R. Golub T.R. MicroRNA expression profiles classify human cancers.Nature. 2005; 435: 834-838Crossref PubMed Scopus (8201) Google Scholar Volinia et al,12Volinia S. Calin G.A. Liu C.G. Ambs S. Cimmino A. Petrocca F. Visone R. Iorio M. Roldo C. Ferracin M. Prueitt R.L. Yanaihara N. Lanza G. Scarpa A. Vecchione A. Negrini M. Harris C.C. Croce C.M. A microRNA expression signature of human solid tumors defines cancer gene targets.Proc Natl Acad Sci U S A. 2006; 103: 2257-2261Crossref PubMed Scopus (4942) Google Scholar and work from Rosetta Genomics13Meiri E. Mueller W.C. Rosenwald S. Zepeniuk M. Klinke E. Edmonston T.B. Werner M. Lass U. Barshack I. Feinmesser M. Huszar M. Fogt F. Ashkenazi K. Sanden M. Goren E. Dromi N. Zion O. Burnstein I. Chajut A. Spector Y. Aharonov R. A second-generation microRNA-based assay for diagnosing tumor tissue origin.Oncologist. 2012; 17: 801-812Crossref PubMed Scopus (119) Google Scholar, 14Pentheroudakis G. Pavlidis N. Fountzilas G. Krikelis D. Goussia A. Stoyianni A. Sanden M. St Croix B. Yerushalmi N. Benjamin H. Meiri E. Chajut A. Rosenwald S. Aharonov R. Spector Y. Novel microRNA-based assay demonstrates 92% agreement with diagnosis based on clinicopathologic and management data in a cohort of patients with carcinoma of unknown primary.Mol Cancer. 2013; 12: 57Crossref PubMed Scopus (37) Google Scholar, 15Rosenfeld N. Aharonov R. Meiri E. Rosenwald S. Spector Y. Zepeniuk M. Benjamin H. Shabes N. Tabak S. Levy A. Lebanony D. Goren Y. Silberschein E. Targan N. Ben-Ari A. Gilad S. Sion-Vardy N. Tobar A. Feinmesser M. Kharenko O. Nativ O. Nass D. Perelman M. Yosepovich A. Shalmon B. Polak-Charcon S. Fridman E. Avniel A. Bentwich I. Bentwich Z. Cohen D. Chajut A. Barshack I. MicroRNAs accurately identify cancer tissue origin.Nature Biotechnol. 2008; 26: 462-469Crossref PubMed Scopus (858) Google Scholar, 16Rosenwald S. Gilad S. Benjamin S. Lebanony D. Dromi N. Faerman A. Benjamin H. Tamir R. Ezagouri M. Goren E. Barshack I. Nass D. Tobar A. Feinmesser M. Rosenfeld N. Leizerman I. Ashkenazi K. Spector Y. Chajut A. Aharonov R. Validation of a microRNA-based qRT-PCR test for accurate identification of tumor tissue origin.Mod Pathol. 2010; 23: 814-823Crossref PubMed Scopus (109) Google Scholar and was recently reviewed by Di Leva and Croce.17Di Leva G. Croce C.M. miRNA profiling of cancer.Curr Opin Genet Dev. 2013; 23: 3-11Crossref PubMed Scopus (344) Google Scholar Besides their tissue specificity, a main advantage of miRNAs as biomarkers is their short size, which renders them more stable in formalin-fixed, paraffin-embedded (FFPE) material compared with mRNA.18Liu A. Tetzlaff M.T. Vanbelle P. Elder D. Feldman M. Tobias J.W. Sepulveda A.R. Xu X. MicroRNA expression profiling outperforms mRNA expression profiling in formalin-fixed paraffin-embedded tissues.Int J Clin Exp Pathol. 2009; 2: 519-527PubMed Google Scholar, 19Siebolts U. Varnholt H. Drebber U. Dienes H.P. Wickenhauser C. Odenthal M. Tissues from routine pathology archives are suitable for microRNA analyses by quantitative PCR.J Clin Pathol. 2009; 62: 84-88Crossref PubMed Scopus (91) Google Scholar By applying a microarray platform based on locked nucleic acid (LNA)-modified detection probes,20Castoldi M. Benes V. Hentze M.W. Muckenthaler M.U. miChip: a microarray platform for expression profiling of microRNAs based on locked nucleic acid (LNA) oligonucleotide capture probes.Methods. 2007; 43: 146-152Crossref PubMed Scopus (68) Google Scholar which enable highly sensitive and specific detection of >2000 miRNAs, we identified tissue-specific miRNA signatures for 35 tumors and histologies, of which 15 were selected for classification. In this study, we evaluate the potential of miRNA expression profiling to identify the primary tumor in patients with cancer. To this end, we have developed a multiclass classification algorithm, which can identify the site of tumor origin with high specificity on the basis of the miRNA profile of the metastasis. We here describe the development of this classifier, which is based on a comprehensive miRNA expression data set. More than 1100 FFPE tumor (both primary and metastases) and normal adjacent tissue samples were procured from the National Disease Research Interchange (Philadelphia, PA), Cytomyx (Lexington, MA), Proteogenex (Culver City, CA), and our in-house tissue bank. Every sample was obtained with a copy of its anonymized pathological report, and both the pathology information and an H&E section of each preparation was reviewed by a pathologist (A.H.) to ascertain the diagnosis, origin, and tumor percentage of the sample. Inclusion criteria for subsequent RNA extraction and miRNA expression analysis were >0.5-mm2 tumor size, <25% normal adjacent tissue, <20% necrosis or hemorrhage, and confirmed histology. In the pilot phase of the project, we collected 408 samples from 35 different tumor histologies to cover a broad selection of solid tumors, whereas for the classifier, we narrowed down the list of included tissues to 15, to represent only the clinically most relevant histologies to identify tumors of unknown origin (Table 1). All demographic metadata were deposited in a database and are available in Supplemental Table S1. For validation of the classifier, an independent set of 48 metastases with known origin was collected from the National Disease Research Interchange and our in-house tissue-bank.Table 1Number of Samples per Tissue, TP, Mean PPV, and Sensitivity (with CIs) of the Classification, Assessed by Fivefold Cross-Validation of the ClassifierTissueHistologySamples (n)TPMean PPV (%)Mean sensitivity (%)AdrenalACC8610075 (41–93)Bile ductCholangiocarcinoma181410078 (55–91)ColorectalAdenocarcinoma, mucinous adenocarcinoma17137776 (53–90)EG junction∗The EG junction class combines samples from esophagus and gastric cancers.Adenocarcinoma, signet cell, mucinous adenocarcinoma, (squamous excluded)20178385 (64–95)Germ cell tumorNonseminoma, seminoma, embryonal carcinoma, yolk sac carcinoma7783100 (65–100)GIST†For some tissue types, the number of samples is relatively low; therefore, the validation results for these histologies should be interpreted with caution.Gastrointestinal stromal tumor5410080 (38–99)KidneyPapillary cell carcinoma, clear cell carcinoma20188790 (70–97)LungAdenocarcinoma (squamous excluded)20188690 (70–97)LymphomaB cell, large cell, marginal zone Hodgkin’s13129593 (67–100)MelanomaMalignant melanoma99100100 (70–100)OvarySerous, mucinous, endometrioid adenocarcinoma, clear cell20139065 (43–82)PancreasDuctal adenocarcinoma, mucinous noncystic20168080 (58–92)Prostate†For some tissue types, the number of samples is relatively low; therefore, the validation results for these histologies should be interpreted with caution.Adenocarcinoma5410080 (38–99)Thyroid†For some tissue types, the number of samples is relatively low; therefore, the validation results for these histologies should be interpreted with caution.Papillary, Hürthle cell, follicular carcinoma66100100 (61–100)Urinary bladderTransitional cell carcinoma, papillary and nonpapillary20198395 (76–100)Total208176ACC, adrenal cortical carcinoma; EG, esophagogastric; PPV, positive predictive value; TP, true positive count.∗ The EG junction class combines samples from esophagus and gastric cancers.† For some tissue types, the number of samples is relatively low; therefore, the validation results for these histologies should be interpreted with caution. Open table in a new tab ACC, adrenal cortical carcinoma; EG, esophagogastric; PPV, positive predictive value; TP, true positive count. Total RNA was extracted from 20-μm FFPE sections with the High Pure miRNA Isolation Kit (Roche Applied Science, Mannheim, Germany) according to the manufacturer’s instructions. After elution in 40 μL of RNase free water, the RNA concentration (A260 nm) and purity (A260/280 and A260/230 ratios) were assessed with a Nanodrop ND-1000 spectrophotometer (Thermo Scientific, Wilmington DE). The RNA was stored at −80°C until further analysis. For microarray analysis, we applied a common reference design in which the reference sample contains a mixture of total RNA to represent all tissue types in the study. This allows for both one- and two-channel data analysis, as described in detail by Søkilde et al.21Søkilde R. Kaczkowski B. Barken K. Mouritzen P. Møller S. Litman T. MicroRNA expression analysis by LNA enhanced microarrays.in: Gusev Y. MicroRNA Profiling in Cancer: A Bioinformatics Perspective. Pan Stanford Publishing, Singapore2009: 23-46Crossref Google Scholar In the present study, we applied the two-channel ratio analysis, because this permits comparison across different array versions. One microgram of total RNA from each sample was labeled by using the miRCURY LNA microRNA Power labeling Kit (Exiqon, Vedbæk, Denmark), according to a two-step protocol as follows: calf intestinal alkaline phosphatase was applied to remove terminal 5′ phosphates, and fluorescent labels were attached enzymatically to the 3′ end of the miRNAs. Sample-specific RNA was labeled with Hy3 (green) fluorophore, whereas the common reference RNA pool was labeled with the Hy5 (red). The Hy3- and Hy5-labeled RNA samples were mixed and co-hybridized to miRCURY LNA Arrays version Dx10 and version 11 (Exiqon), which contain Tm-normalized capture probes that target miRNAs from human, mouse, and rat, as registered in miRBase version 19.0 at the Sanger Institute.22Kozomara A. Griffiths-Jones S. miRBase: integrating microRNA annotation and deep-sequencing data.Nucleic Acids Res. 2011; 39: D152-D157Crossref PubMed Scopus (2948) Google Scholar Hybridization was performed overnight for 16 hours at 65°C in a Tecan HS4800 hybridization station (Tecan, Männedorf, Switzerland). After washing and drying, the microarray slides were scanned under ozone-free conditions (ozone level < 2.0 ppb to minimize bleaching of the fluorescent dyes) in a G2565BA Microarray Scanner System (Agilent, Santa Clara, CA). The resulting images were quantified with Imagene software version 8.0 (BioDiscovery, El Segundo, CA), and both automatic quality control (flagging of poor spots by the software) and manual, visual inspection were performed to ensure the highest possible data quality. The expression levels of 39 selected miRNAs were validated by quantitative real-time PCR to apply the miRCURY LNA Universal RT microRNA PCR system and SYBR Green master mix according to the manufacturer’s instructions (Exiqon). The results are shown in Supplemental Figure S1. All low-level analyses were performed in the R environment, including importing and preprocessing of the data with the use of the LIMMA package (http://www.bioconductor.org/packages/2.13/bioc/html/limma.html, last accessed August 29, 2013). Mean pixel intensities were used to calculate signal (foreground) spot intensities, and median pixel intensities were applied to estimate background intensity. After excluding flagged spots from the analysis, the normexp background correction method, with offset equal to 10, was applied.23Ritchie M.E. Silver J. Oshlack A. Holmes M. Diyagama D. Holloway A. Smyth G.K. A comparison of background correction methods for two-colour microarrays.Bioinformatics. 2007; 23: 2700-2707Crossref PubMed Scopus (730) Google Scholar For intraslide normalization, the global Lowess (Locally Weighted Scatterplot Smoothing) regression algorithm was applied, and log2 ratios of four intraslide replicates were averaged. All expression data were deposited in the Rosetta Resolver (Rosetta Biosoftware, Hoddesdon, UK) data management and analysis system. A miRNA expression database was built to identify miRNAs with high discriminatory power between tumor histologies. Three approaches for feature selection (that is, filters, wrappers, and embedded methods) are commonly used.24Ma S. Huang J. Penalized feature selection and classification in bioinformatics.Brief Bioinform. 2008; 9: 392-403Crossref PubMed Scopus (177) Google Scholar Here, we have applied both filtering and a wrapper; differentially expressed miRNAs were identified by running a one versus one, as well as a one versus all t-tests for each histology, followed by ranking of the most significant candidate miRNAs. In addition, the feature selection embedded in the least absolute shrinkage and selection operator (LASSO) classification algorithm was applied. The LASSO classifier was originally described by Tibshirani25Tibshirani R. Regression shrinkage and selection via the lasso.J Royal Stat Soc. 1996; 58: 267-288Google Scholar and is based on a multinomial logistic model, which is fitted by using L1 regularization.26Efron B. Hastie T. Johnstone I. Tibshirani R. Least angle regression.Ann Statist. 2004; 32: 409-499Google Scholar The regularization parameter is chosen by evaluating the results of a cross-validation along the entire regularization path. To solve the L1 regularized optimization problem we used the glmnet algorithm.27Friedman J. Hastie T. Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent.J Stat Softw. 2010; 33: 1-22Crossref PubMed Scopus (8749) Google Scholar The classifier was built on log2 ratio data from the 208 samples and 15 cancer classes listed in Table 1. We tested and fivefold cross-validated the LASSO algorithm and have listed its model coefficient, a measure of discriminatory potential, in Supplemental Table S2. For the present multiclass classification task, we found that LASSO performed on par with or even better than other classification algorithms, such as K nearest neighbor and linear discriminant analysis (data not shown). All calculations and statistical tests were done in the free software environment for statistical computing and graphics R version 2.9.2 (http://www.r-project.org, last accessed August 29, 2013). For microarray analysis, the open source package for R, Bioconductor, was used (http://www.bioconductor.org). Confidence intervals were calculated with the Wilson method by using the R binom library, and the following script: binom.confint(x, n, conf.level = 0.95, methods = wilson), where x = number of successes and n = number of independent trials. To obtain as comprehensive a data set as possible for constructing the microarray tumor database, we initially profiled 1129 samples that spanned most tumor sites and covered 35 major histological subtypes. When considering which tissue classes to include in the final classifier, we focused on those metastatic cancers that are most frequently found, that is, at autopsy, in CUP origin. Greater than 75% of all CUP cases are adenocarcinomas and poorly differentiated carcinomas, of which the most common primary sites (when determined) are pancreas (25%), lung (20%), stomach, colorectum, and hepatobiliary tract (8% to 12% each), and kidney (5%). Squamous cell carcinomas account for 10% to 15%, most of which arise from head and neck tumors, whereas melanoma represents 4% of all CUP cases. These relative frequencies, however, should be interpreted with caution, because the epidemiology of CUP is changing due to both improved medical imaging technology and lifestyle habits; therefore, different studies report dissimilar frequencies of primary sites.28Pentheroudakis G. Golfinopoulos V. Pavlidis N. Switching benchmarks in cancer of unknown primary: from autopsy to microarray.Eur J Cancer. 2007; 43: 2026-2036Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar On the basis of the above considerations, our selection of tissues includes the major carcinoma (12 of 15 histologies), as well as melanoma, germ cell tumors (clear cell tumors), and lymphoma (small cell neoplasms), because these can be difficult to distinguish from poorly differentiated carcinoma. Finally, taking into account that in the clinical setting FFPE material is readily available and, thus, represents an important resource for molecular profiling, and that miRNAs are stable in FFPE blocks and straightforward to extract,29Li J. Smyth P. Flavin R. Cahill S. Denning K. Aherne S. Guenther S.M. O’Leary J.J. Sheils O. Comparison of miRNA expression patterns using total RNA extracted from matched samples of formalin-fixed paraffin-embedded (FFPE) cells and snap frozen cells.BMC Biotechnol. 2007; 7: 36Crossref PubMed Scopus (297) Google Scholar we decided to develop the classifier on FFPE material. Table 1 lists the 15 tissues and histologies (columns 1 and 2), which were included in the training set that consisted of 208 FFPE samples (199 primary tumors and 9 metastases). A detailed summary of all patient demographic data can be found in Supplemental Table S1, and the expression data are deposited in Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo; accession number GSE50894). The distribution of tissue-specific miRNAs (ie, those miRNAs that were preferentially expressed in samples originating from one tissue compared with all other tissues) is summarized in the heatmap (Figure 1). From the heatmap it is evident that some histologies are easy to distinguish from the rest because of a strong and homogeneous tissue-specific miRNA signature [adrenal, lymphoma, germ cell, prostate, gastrointestinal stromal tumor (GIST), and melanoma], whereas other tissue origins are more difficult to classify accurately, mainly because of heterogeneity within the group (ovary, lung) or because of high similarity to related tissue types [colorectal and esophagogastric (EG) junction]. Because selection of the candidate biomarkers is crucial for performance of the classifier, we took several different approaches to identify the best possible tissue-specific markers. The first and simplest approach was to run one-against-one and one-against-all comparisons for each tissue, identifying differentially expressed miRNAs by t-tests. However, because running multiple two-sample t-tests can result in an increased risk of committing a type I error (false positive), we also applied analysis of variance to compare all 15 means (of the different histologies) in one test. Yet, because filtering-based methods, such as t-test and analysis of variance, do not provide a cross-validation option for optimization of the set of discriminatory features, we decided for an embedded approach, namely the LASSO method, which integrates feature selection within the classifier construction. With this method 132 miRNAs with high tissue discriminatory potential were identified; these are listed in Supplemental Table S2, which is a data matrix showing each feature’s LASSO model coefficient for the particular tissue of interest. Finally, we made a literature search for tissue-specific miRNAs and compared these with our top candidate discriminatory miRNAs. There was, not surprisingly, a high degree of overlap between the miRNAs identified in our study and those reported previously as having high predictive ability for cancer classification.12Volinia S. Calin G.A. Liu C.G. Ambs S. Cimmino A. Petrocca F. Visone R. Iorio M. Roldo C. Ferracin M. Prueitt R.L. Yanaihara N. Lanza G. Scarpa A. Vecchione A. Negrini M. Harris C.C. Croce C.M. A microRNA expression signature of human solid tumors defines cancer gene targets.Proc Natl Acad Sci U S A. 2006; 103: 2257-2261Crossref PubMed Scopus (4942) Google Scholar, 15Rosenfeld N. Aharonov R. Meiri E. Rosenwald S. Spector Y. Zepeniuk M. Benjamin H. Shabes N. Tabak S. Levy A. Lebanony D. Goren Y. Silberschein E. Targan N. Ben-Ari A. Gilad S. Sion-Vardy N. Tobar A. Feinmesser M. Kharenko O. Nativ O. Nass D. Perelman M. Yosepovich A. Shalmon B. Polak-Charcon S. Fridman E. Avniel A. Bentwich I. Bentwich Z. Cohen D. Chajut A. Barshack I. MicroRNAs accurately identify cancer tissue origin.Nature Biotechnol. 2008; 26: 462-469Crossref PubMed Scopus (858) Google Scholar, 16Rosenwald S. Gilad S. Benjamin S. Lebanony D. Dromi N. Faerman A. Benjamin H. Tamir R. Ezagouri M. Goren E. Barshack I. Nass D. Tobar A. Feinmesser M. Rosenfeld N. Leizerman I. Ashkenazi K. Spector Y. Chajut A. Aharonov R. Validation of a microRNA-based qRT-PCR test for accurate identification of tumor tissue origin.Mod Pathol. 2010; 23: 814-823Crossref PubMed Scopus (109) Google Scholar The overlapping miRNAs are also indicated in Supplemental Table S2. The miRNAs that can be used for classification of tumor origin are listed in Table 2.Table 2The miRNAs That Can Be Used for Identification of Tumor OriginTissuemiRNAACChsa-miR-129∗, hsa-miR-136, hsa-miR-202∗, hsa-miR-218, hsa-miR-376c, hsa-miR-488Bile duct; cholangiocarcinomahsa-miR-23a, hsa-miR-122, hsa-miR-214, hsa-miR-452, hsa-miR-616Colorectal; adenocarcinoma, mucinous adenocarcinomahsa-miR-26b∗, hsa-miR-95, hsa-miR-99b∗ hsa-miR-134, hsa-miR-192∗, hsa-miR-194 hsa-miR-196b, hsa-miR-220b, hsa-miR-224 hsa-miR-433, hsa-miR-491-5p, hsa-miR-516a-3p hsa-miR-629∗, hsa-miR-767-3p, hsa-miR-890EG junction; adenocarcinoma, signet cell, mucinous adenocarcinoma (squamous excluded)†The EG junction class combines samples from esophagus and gastric cancers.hsa-miR-7, hsa-miR-16-1∗, hsa-miR-96∗ hsa-miR-124, hsa-miR-133b, hsa-miR-143 hsa-miR-145∗, hsa-miR-147b, hsa-miR-450b-3p‡miR-323 was previously named miR-453. hsa-miR-323, hsa-miR-504, hsa-miR-548a-3p hsa-miR-548b-5p, hsa-miR-647, hsa-miR-892bGerm cell tumor; nonseminoma, seminoma, embryonal carcinoma, yolk sac carcinomahsa-miR-154∗, hsa-miR-367, hsa-miR-372 hsa-miR-423-3p, hsa-miR-769-3pGISThsa-miR-132, hsa-miR-574-3p, hsa-miR-603Kidney; papillary cell carcinoma, clear cell carcinomahsa-miR-10b, hsa-miR-30a∗, hsa-miR-92a-1∗ hsa-miR-105, hsa-miR-148a∗, hsa-miR-196a hsa-miR-199b-5p, hsa-miR-204, hsa-miR-210 hsa-miR-340, hsa-miR-491-3p, hsa-miR-557Lung; adenocarcinoma (squamous excluded)hsa-miR-23a∗, hsa-miR-34b∗, hsa-miR-34c-5p hsa-miR-96, hsa-miR-126∗, hsa-miR-129-3p hsa-miR-185, hsa-miR-193b, hsa-miR-212 hsa-miR-217, hsa-miR-219-5p, hsa-miR-601Lymphoma; B cell, large cell, marginal zone Hodgkin’shsa-miR-10a, hsa-miR-27b, hsa-miR-142-5p hsa-miR-153, hsa-miR-155, hsa-miR-155∗ hsa-miR-451, hsa-miR-541∗, hsa-miR-615-5p hsa-miR-641Melanomahsa-miR-146a, hsa-miR-150∗, hsa-miR-211 hsa-miR-541∗Ovary; serous, mucinous, endometrioid adenocarcinoma, clear cellhsa-miR-92b, hsa-miR-130a, hsa-miR-130a∗ hsa-miR-135a, hsa-miR-141, hsa-miR-142-3p hsa-miR-330-5p, hsa-miR-499-5p, hsa-miR-514 hsa-miR-519c-3p, hsa-miR-522, hsa-miR-572 hsa-miR-592, hsa-miR-708, hsa-miR-923Pancreas; ductal adenocarcinoma, mucinous noncystichsa-miR-199a-3p, hsa-miR-221∗, hsa-miR-335 hsa-miR-431∗, hsa-miR-454∗, hsa-miR-582-3p hsa-" @default.
• W2095194002 created "2016-06-24" @default.
• W2095194002 creator A5010004397 @default.
• W2095194002 creator A5040281361 @default.
• W2095194002 creator A5048363162 @default.
• W2095194002 creator A5052400382 @default.
• W2095194002 creator A5061533768 @default.
• W2095194002 creator A5067507410 @default.
• W2095194002 creator A5081329468 @default.
• W2095194002 creator A5082453853 @default.
• W2095194002 creator A5084556694 @default.
• W2095194002 creator A5090656813 @default.
• W2095194002 date "2014-01-01" @default.
• W2095194002 modified "2023-10-18" @default.
• W2095194002 title "Efficient Identification of miRNAs for Classification of Tumor Origin" @default.
• W2095194002 cites W1648301895 @default.
• W2095194002 cites W1970488378 @default.
• W2095194002 cites W1975410472 @default.
• W2095194002 cites W1979166355 @default.
• W2095194002 cites W1981272958 @default.
• W2095194002 cites W1986756287 @default.
• W2095194002 cites W1987930055 @default.
• W2095194002 cites W1989688154 @default.
• W2095194002 cites W1996760332 @default.
• W2095194002 cites W2014427523 @default.
• W2095194002 cites W2016525674 @default.
• W2095194002 cites W2017426710 @default.
• W2095194002 cites W2025541701 @default.
• W2095194002 cites W2025731839 @default.
• W2095194002 cites W2029670310 @default.
• W2095194002 cites W2036273883 @default.
• W2095194002 cites W2044337456 @default.
• W2095194002 cites W2063978378 @default.
• W2095194002 cites W2075986328 @default.
• W2095194002 cites W2087797727 @default.
• W2095194002 cites W2109363337 @default.
• W2095194002 cites W2123280474 @default.
• W2095194002 cites W2124014074 @default.
• W2095194002 cites W2130979840 @default.
• W2095194002 cites W2135810600 @default.
• W2095194002 cites W2136159021 @default.
• W2095194002 cites W2137317896 @default.
• W2095194002 cites W2137476312 @default.
• W2095194002 cites W2138687531 @default.
• W2095194002 cites W2141670106 @default.
• W2095194002 cites W2144149602 @default.
• W2095194002 cites W2146519879 @default.
• W2095194002 cites W2149106795 @default.
• W2095194002 cites W2162359611 @default.
• W2095194002 cites W2164216268 @default.
• W2095194002 cites W2166074036 @default.
• W2095194002 cites W2168230285 @default.
• W2095194002 cites W2171697766 @default.
• W2095194002 cites W2172241845 @default.
• W2095194002 cites W4231147183 @default.
• W2095194002 cites W4256123302 @default.
• W2095194002 cites W4294541781 @default.
• W2095194002 doi "https://doi.org/10.1016/j.jmoldx.2013.10.001" @default.
• W2095194002 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/24211363" @default.
• W2095194002 hasPublicationYear "2014" @default.
• W2095194002 type Work @default.
• W2095194002 sameAs 2095194002 @default.
• W2095194002 citedByCount "48" @default.
• W2095194002 countsByYear W20951940022014 @default.
• W2095194002 countsByYear W20951940022015 @default.
• W2095194002 countsByYear W20951940022016 @default.
• W2095194002 countsByYear W20951940022017 @default.
• W2095194002 countsByYear W20951940022018 @default.
• W2095194002 countsByYear W20951940022019 @default.
• W2095194002 countsByYear W20951940022020 @default.
• W2095194002 countsByYear W20951940022021 @default.
• W2095194002 countsByYear W20951940022022 @default.
• W2095194002 countsByYear W20951940022023 @default.
• W2095194002 crossrefType "journal-article" @default.
• W2095194002 hasAuthorship W2095194002A5010004397 @default.
• W2095194002 hasAuthorship W2095194002A5040281361 @default.
• W2095194002 hasAuthorship W2095194002A5048363162 @default.
• W2095194002 hasAuthorship W2095194002A5052400382 @default.
• W2095194002 hasAuthorship W2095194002A5061533768 @default.
• W2095194002 hasAuthorship W2095194002A5067507410 @default.
• W2095194002 hasAuthorship W2095194002A5081329468 @default.
• W2095194002 hasAuthorship W2095194002A5082453853 @default.
• W2095194002 hasAuthorship W2095194002A5084556694 @default.
• W2095194002 hasAuthorship W2095194002A5090656813 @default.
• W2095194002 hasBestOaLocation W20951940021 @default.
• W2095194002 hasConcept C104317684 @default.
• W2095194002 hasConcept C116834253 @default.
• W2095194002 hasConcept C145059251 @default.
• W2095194002 hasConcept C54355233 @default.
• W2095194002 hasConcept C59822182 @default.
• W2095194002 hasConcept C70721500 @default.
• W2095194002 hasConcept C86803240 @default.
• W2095194002 hasConceptScore W2095194002C104317684 @default.
• W2095194002 hasConceptScore W2095194002C116834253 @default.
• W2095194002 hasConceptScore W2095194002C145059251 @default.
• W2095194002 hasConceptScore W2095194002C54355233 @default.
• W2095194002 hasConceptScore W2095194002C59822182 @default.
• W2095194002 hasConceptScore W2095194002C70721500 @default.

• next