Development and validation of a NanoString BASE47 bladder cancer gene classifier
Ontology highlight
ABSTRACT:
ORGANISM(S): Homo sapiens
PROVIDER: GSE160693 | GEO | 2020/11/03
REPOSITORIES: GEO
Similar Datasets
Project description:The optimal treatment of patients with cancer depends on establishing accurate diagnoses by using a complex combination of clinical and histopathological data. In some instances, this task is difficult or impossible because of atypical clinical presentation or histopathology. To determine whether the diagnosis of multiple common adult malignancies could be achieved purely by molecular classification, we subjected 218 tumor samples, spanning 14 common tumor types, and 90 normal tissue samples to oligonucleotide microarray gene expression analysis. The expression levels of 16,063 genes and expressed sequence tags were used to evaluate the accuracy of a multiclass classifier based on a support vector machine algorithm. Overall classification accuracy was 78%, far exceeding the accuracy of random classification (9%). Poorly differentiated cancers resulted in low-confidence predictions and could not be accurately classified according to their tissue of origin, indicating that they are molecularly distinct entities with dramatically different gene expression patterns compared with their well differentiated counterparts. Taken together, these results demonstrate the feasibility of accurate, multiclass molecular cancer classification and suggest a strategy for future clinical implementation of molecular cancer diagnostics.
2015-05-16 | GSE68928 | GEO
Project description:Pediatric acute lymphoblastic leukemia (ALL) contains cytogenetically distinct subtypes that respond differently to cytotoxic drugs. Subtype classification can be also achieved through gene expression profiling. However, how to apply such classifiers to a single patient and correctly diagnose the disease subtype in an independent patient group has not been addressed. Furthermore, the underlying regulatory mechanisms responsible for the subtype-specific gene expression patterns are still largely unknown. Here, by combining three published microarray datasets (PMIDs: 12086872, 12730115, 17002788) on 535 Caucasian samples and generating a new dataset on 100 Chinese children ALL samples, we were able to 1) identify a 62-gene classifier with 97.6% accuracy from the Caucasian samples and validated it on the completely independent set of 100 Chinese samples, 2) to uncover potential regulatory networks of ALL subtypes. The classifier we identified was so far the only one that could be applied directly to a single sample and sustained validation in a large independent patient group. Our results also suggest that the etiology of ALL is largely the same among different ethnic groups, and that the transcription factor hubs in the predicted regulatory network might play important roles in regulating gene expression and development of ALL. A total of 100 Chinese pediatric acute lymphoblastic leukemia bone marrow (BM) samples were analyzed, together with five non-ALL BM samples as negative control (the five samples were combined to generate one control, see details in Li et al. Blood, 2009). The diagnosis of ALL was based on morphology, immunology, cytogenetic and molecular (MICM) classification. Cytogenetic ALL subtypes were identified experimentally by G-banding karyotype and multiplex nested RT-PCR. Among the 100 ALL patients, eleven relapsed within five years. All the samples, including those relapsed afterward, were from patients treated on BCH-2003 protocol and were extracted at their initial diagnosis. The five non-ALL BM samples were taken from the removed bones of patients who had plastic surgery for their bone deformity in Beijing Children’s Hospital. And the informed consent was obtained from parents, guardians, or patients (as appropriate). More details are available in Li et al. Gene expression-based classification and regulatory networks of pediatric acute lymphoblastic leukemia, Blood, 2009. We have generated these ALL and control samples to validate that a high accuracy subtype classifier that we identified could be applied directly to a single sample and sustained validation in a large independent patient group.
2009-08-19 | E-GEOD-17703 | biostudies-arrayexpress
Project description:Our aim to develop the 95GC (gene classifier) was to make an accurate diagnostic system using gene expression analysis by means of DNA microarray for prognosis of node-negative and estrogen receptor (ER)-positive breast cancer patients in order to identify a subset of patients who can be safely spared adjuvant chemotherapy. A diagnostic system comprising a 95-gene classifier was developed for predicting the prognosis of node-negative and ER-positive breast cancer patients by using DNA microarray (gene expression) data (n = 549) as the training set and the DNA microarray data (n = 105) as the validation set (= this data set). With the 95-gene classifier we could classify the 105 patients receiving only endocrine therapy without chemotherapy in the validation set into a high-risk (n = 44) and a low-risk (n = 61) group with 10-year recurrence-free survival rates of 93 and 53%, respectively (P = 8.6e-7). Multivariate analysis demonstrated that the 95-gene classifier was the most important and significant predictor of recurrence (P = 9.6e-4) independently of tumor size, histological grade, progesterone receptor, HER2, Ki67, or GGI. The 95-gene classifier developed by us can predict the prognosis of node-negative and ER-positive breast cancer patients with high accuracy. *Note: This old data has been updated multiple times by others. Then, there are some differences from the original 2011 paper and unclear points still remain. Therefore, do not use it for formal analysis aimed at public insurance coverage etc. This is for research purposes only. Please cite this paper when writing a new paper. DOI: 10.1007/s10549-010-1145-z
2023-05-28 | GSE233251 | GEO
Project description:BACKGROUND: In patients with suspicious pulmonary lesions, bronchoscopy is frequently non-diagnostic. This often results in additional invasive testing, including surgical biopsy, although many patients have benign disease. We sought to validate an airway gene-expression classifier for lung cancer in patients undergoing diagnostic bronchoscopy. METHODS: Two multicenter prospective studies (AEGIS 1 and 2) enrolled 1357 current or former smokers undergoing bronchoscopy for suspected lung cancer. Bronchial epithelial cells were collected from normal appearing mucosa in the mainstem bronchus during bronchoscopy. Patients without a definitive diagnosis from bronchoscopy were followed for 12 months. A gene-expression classifier was used to assess the risk of lung cancer, and its performance was evaluated. RESULTS: A total of 298 patients from AEGIS 1 and 341 from AEGIS 2 met criteria for analysis. Bronchoscopy was non-diagnostic for cancer in 272 of 639 patients (43%; 95%CI, 39-46%). The gene expression classifier correctly identified 431 of 487 patients with cancer (89% sensitivity; 95%CI, 85-91%), and 72 of 152 patients without cancer (47% specificity; 95%CI, 40-55%). The combination of the classifier and bronchoscopy had a sensitivity of 97% (95%CI, 95-98%), which was independent of size, location, stage, and histological subtype of lung cancer. In patients with an intermediate pre-test risk (10-60%) of lung cancer, the NPV of the classifier was 91% (95%CI 75-98%). CONCLUSIONS: In patients with an intermediate risk of lung cancer and a non-diagnostic bronchoscopy, a gene-expression classification of “low-risk” warrants consideration of a more conservative diagnostic approach that could reduce unnecessary invasive testing in patients with benign disease. 680 CEL files from 639 BEC specimens, 152 benign and 487 malignant samples
2015-05-17 | E-GEOD-66499 | biostudies-arrayexpress
Project description:The optimal treatment of patients with cancer depends on establishing accurate diagnoses by using a complex combination of clinical and histopathological data. In some instances, this task is difficult or impossible because of atypical clinical presentation or histopathology. To determine whether the diagnosis of multiple common adult malignancies could be achieved purely by molecular classification, we subjected 218 tumor samples, spanning 14 common tumor types, and 90 normal tissue samples to oligonucleotide microarray gene expression analysis. The expression levels of 16,063 genes and expressed sequence tags were used to evaluate the accuracy of a multiclass classifier based on a support vector machine algorithm. Overall classification accuracy was 78%, far exceeding the accuracy of random classification (9%). Poorly differentiated cancers resulted in low-confidence predictions and could not be accurately classified according to their tissue of origin, indicating that they are molecularly distinct entities with dramatically different gene expression patterns compared with their well differentiated counterparts. Taken together, these results demonstrate the feasibility of accurate, multiclass molecular cancer classification and suggest a strategy for future clinical implementation of molecular cancer diagnostics. golub-00236 Assay Type: Gene Expression Provider: Affymetrix Array Designs: Hu35KsubA, Hu6800 Organism: Homo sapiens (ncbitax) Material Types: synthetic_RNA, organism_part, whole_organism, total_RNA Disease States: Colorectal Adenocarcinoma, Melanoma, Normal, pancreatic adenocarcinoma, Acute Myeloid Leukemia, Breast Adenocarcinoma, T-cell ALL, Lung Adenocarcinoma, bladder transitional cell carcinoma, Ovarian Adenocarcinoma, Mesothelioma, B-cell ALL, prostate adenocarcinoma, Medulloblastoma, Follicular Lymphoma, renal cell carcinoma, uterine adenocarcinoma, large B-cell lymphoma, Glioblastoma, bladder transitonal cell carcinoma
2015-05-16 | E-GEOD-68928 | biostudies-arrayexpress
Project description:The analytes qualified as biomarkers are potent tools to diagnose various diseases, monitor therapy responses, and design therapeutic interventions. The early assessment of the diverseness of human disease is essential for the speedy and cost-efficient implementation of personalized medicine. We developed g3mclass, the Gaussian mixture modeling software for molecular assay data classification. This software automates the validated multiclass classifier applicable to single analyte tests and multiplexing assays. The g3mclass achieves automation using the original semi-constrained expectation-maximization (EM) algorithm that allows inference from the test, control, and query data that human experts cannot interpret. In this study, we used real-world clinical data and gene expression datasets (ERBB2, ESR1, PGR) to provide examples of how g3mclass may help overcome the problems of over-/underdiagnosis and equivocal results in diagnostic tests for breast cancer. We showed the g3mclass output’s accuracy, robustness, scalability, and interpretability. The user-friendly interface and free dissemination of this multi-platform software aim to ease its use by research laboratories, biomedical pharma, companion diagnostic developers, and healthcare regulators. Furthermore, the g3mclass automatic extracting information through probabilistic modeling is adaptable for blending with machine learning and artificial intelligence.
2022-11-09 | GSE214540 | GEO
Project description:Many innovative techniques and scientific improvements are available to tackle societal concerns, like public health safety and confining the risk of cancerous exposure to chemicals, but have not been thoroughly tested and implicated yet. We investigated if microRNA and mRNA transcription profiles can be implemented in a short-term carcinogen classifier assay. Our study is additionally focusing on the drawbacks of present-day carcinogen screening strategies and also aims to contribute to a more ethical approach towards animal use and welfare within risk assessment. Since current in vitro and in silico assays are still not able to mimic the in vivo situation accurately we set out to develop an alternative short-term in vivo assay. Five genotoxic, seven non-genotoxic and five non-carcinogen exposure studies were used to investigate if murine hepatic microRNA and mNA profiles after 7-day exposure are suitable tools to classify carcinogens. Classification analyses showed that a small transcript set, consisting of both microRNA and mRNA, is able to classify the genotoxic, non-genotoxic and non-carcinogens tested with 100% accuracy. The results indicate that microRNAs have the potential to be used as transcriptional classifiers and that a short-term transcriptional classifier assay in mice can be a powerful tool in carcinogenicity risk assessment. Since current in vitro and in silico assays are still not able to mimic the in vivo situation accurately we set out to develop an alternative short-term in vivo assay. Five genotoxic, seven non-genotoxic and five non-carcinogen exposure studies were used to investigate if murine hepatic microRNA and mNA profiles after 7-day exposure are suitable tools to classify carcinogens. Classification analyses showed that a small transcript set, consisting of both microRNA and mRNA, is able to classify the genotoxic, non-genotoxic and non-carcinogens tested with 100% accuracy. The results indicate that microRNAs have the potential to be used as transcriptional classifiers and that a short-term transcriptional classifier assay in mice can be a powerful tool in carcinogenicity risk assessment. [mRNA profling] 96 hepatic samples in total, 8 control untreated samples, replicates per treated group n=4-6
2014-04-01 | E-GEOD-43977 | biostudies-arrayexpress
Project description:Development of a non-invasive, whole blood microarray gene expression signature that accurately distinguishes malignant from benign lung nodules detected by LDCT and the successful transition of that signature to a custom NanoString panel with no loss of accuracy.
2018-12-20 | GSE108375 | GEO
Project description:BACKGROUND: In patients with suspicious pulmonary lesions, bronchoscopy is frequently non-diagnostic. This often results in additional invasive testing, including surgical biopsy, although many patients have benign disease. We sought to validate an airway gene-expression classifier for lung cancer in patients undergoing diagnostic bronchoscopy. METHODS: Two multicenter prospective studies (AEGIS 1 and 2) enrolled 1357 current or former smokers undergoing bronchoscopy for suspected lung cancer. Bronchial epithelial cells were collected from normal appearing mucosa in the mainstem bronchus during bronchoscopy. Patients without a definitive diagnosis from bronchoscopy were followed for 12 months. A gene-expression classifier was used to assess the risk of lung cancer, and its performance was evaluated. RESULTS: A total of 298 patients from AEGIS 1 and 341 from AEGIS 2 met criteria for analysis. Bronchoscopy was non-diagnostic for cancer in 272 of 639 patients (43%; 95%CI, 39-46%). The gene expression classifier correctly identified 431 of 487 patients with cancer (89% sensitivity; 95%CI, 85-91%), and 72 of 152 patients without cancer (47% specificity; 95%CI, 40-55%). The combination of the classifier and bronchoscopy had a sensitivity of 97% (95%CI, 95-98%), which was independent of size, location, stage, and histological subtype of lung cancer. In patients with an intermediate pre-test risk (10-60%) of lung cancer, the NPV of the classifier was 91% (95%CI 75-98%). CONCLUSIONS: In patients with an intermediate risk of lung cancer and a non-diagnostic bronchoscopy, a gene-expression classification of “low-risk” warrants consideration of a more conservative diagnostic approach that could reduce unnecessary invasive testing in patients with benign disease.
2015-05-17 | GSE66499 | GEO
Project description:Contemporary treatment of pediatric acute myeloid leukemia (AML) requires the assignment of patients to specific risk groups. To explore whether expression profiling of leukemic blasts could accurately distinguish between the known risk groups of AML, we analyzed 130 pediatric and 20 adult AML diagnostic bone marrow or peripheral blood samples using the Affymetrix U133A microarray. Class discriminating genes were identified for each of the major prognostic subtypes of pediatric AML, including t(15;17)[PML-RARalpha], t(8;21)[AML1-ETO], inv(16) [CBFbeta-MYH11], MLL chimeric fusion genes, and cases classified as FAB-M7. When subsets of these genes were used in supervised learning algorithms, an overall classification accuracy of more than 93% was achieved. Moreover, we were able to use the expression signatures generated from the pediatric samples to accurately classify adult de novo AMLs with the same genetic lesions. The class discriminating genes also provided novel insights into the molecular pathobiology of these leukemias. Finally, using a combined pediatric data set of 130 AMLs and 137 acute lymphoblastic leukemias, we identified an expression signature for cases with MLL chimeric fusion genes irrespective of lineage. Surprisingly, AMLs containing partial tandem duplications of MLL failed to cluster with MLL chimeric fusion gene cases, suggesting a significant difference in their underlying mechanism of transformation. All the gene expression arrays are available through http://www.stjuderesearch.org/site/data/AML1/ in the original study (PMID:15226186). To study the RAS gene expression in the human AML patients, a total of 104 AML cases with known KRAS and NRAS status (including 72 gene expression arrays in the original study and 32 additional arrays acquired later on), as well as 4 CD34+ normal bone marrow cases deposited in GEO GSE33315, were including in this depository. Gene expression profiling was performed on 104 single diagnosis tumor samples and 4 CD34+ normal bone marrow samples
2013-07-02 | E-GEOD-43176 | biostudies-arrayexpress