Dataset Information

Defining glioma subtypes based on robust transcriptional patterns from 16 prior studies

ABSTRACT: The purpose of our study was to define robust glioma subtypes by applying rigorous preprocessing and validation steps to 1,952 microarray samples aggregated from public data repositories for 16 prior studies. We evaluated each sample for quality-control issues, normalized high-quality samples using the Single-Channel Array Normalization (SCAN) algorithm (PMID: 22959562), corrected for probe-composition biases and inter-platform variability, and adjusted for intra- and inter-study batch effects. The deposited data in GEO include the 1,841 microarray samples that passed quality control tests, and underwent normalization and batch effect adjustment. Where available, we retrieved treatment, histological and clinical data, such as tumor grade, histopathology, age-at-diagnosis, and survival time after diagnosis for these samples. Using a training/testing validation design, we identified six transcriptional subtypes in the training set, and evaluated clinically observable characteristics in the test set. Three of our clusters contained a heterogeneous mix of histopathological subtypes and tumor grades. We evaluated age, survival, and treatment patterns across our test samples and observed highly significant differences among the clusters. We also observed the potential to use gene expression patterns to further understanding of the biological mechanisms that drive gliomagenesis for each subtype. Our findings provide clinical and biological insights that may not be apparent with alternative approaches or smaller data sets, and our approach serves as an example for gene-expression meta-analyses that can be applied to other complex diseases.

ORGANISM(S): Homo sapiens

PROVIDER: GSE55918 | GEO | 2014/12/21

REPOSITORIES: GEO

ACCESS DATA

Dataset's files

Source:

			Action	DRS
		Other

Items per page:

1 - 1 of 1

Similar Datasets

Project description:Lymphomas are classified according to the World Health Organization (WHO) classification which defines subtypes on the basis of clinical, morphological, immunophenotypic, molecular and cytogenetic criteria. Using this model, 8 of 9 of the validation samples were classified successfully. This pilot study demonstrates that such a microarray tool may be a promising diagnostic approach for small B-cell non-Hodgkin’s lymphoma. Keywords: Small B-Cell non-Hodgkin’s Lymphoma, Low Density DNA Microarray, Diagnosis Fresh-frozen tumor biopsies performed before treatment and clinical data were obtained retrospectively from 68 patients in five different institutions after lab investigations involving cytology, immunohistochemistry, cytogenetics (conventional cytogenetics and fluorescent in situ hybridization [FISH]), and/or molecular analysis. The samples were then classified unanimously by a panel of five pathologists as one of the 4 subtypes: 17 B-CLL, 14 MZL, 23 FL and 14 MCL. The immunohistochemical criteria used for identifying the 4 subtypes of SBCL were: CD20+/-, CD5+, CD23+ for B-CLL; CD20+, CD5+, CD23-, Cyclin D1+ for MCL; CD20+, CD10+, Bcl2+ for FL; CD20+, CD5-, CD10- for S-MZL. 28 samples were rejected for technical issues with RNA. The remaining 40 samples include the 4 different subtypes with this following distribution: Follicular Lymphoma/FL (n=15), Mantle Cell Lymphoma/MCL (n=7), B-Chronic Lymphocytic Leukemia/B-CLL (n=6) and Splenic Marginal Zone Lymphoma/S-MZL (n=12). The expression profile of 107 genes was evaluated in these 40 samples defined as a training set. Replicates were performed for samples of sufficient quantity, and a total of 79 gene expression profiles, comprised of 12 samples in triplicate, 15 samples in duplicate and 13 samples evaluated singly, were then analyzed. A second group of 13 patients was defined as test set and used to validate the gene signatures highlighted from the training set analysis. These patients were selected by a pathologist independent from the first analysis and include 4 FL, 3 B-CLL, 5 MCL and 1 S-MZL. Four samples were rejected for technical issues with RNA. The remaining 9 samples include the 4 different subtypes: 2 FL, 3 B-CLL, 3 MCL and 1 S-MZL. A total of 13 gene expression profiles were analyzed (4 samples evaluated in duplicate and 5 samples investigated singly).

Project description:Background and Aims Formalin-fixed, paraffin-embedded (FFPE) tissue is the most commonly available form of archived clinical specimens, which are often stored as thin sections on glass slides. RNA isolated from such archived section (AS) of FFPE tissue is more degraded compared to freshly cut (FC) FFPE section because of prolonged air exposure. In this study, we evaluated performance of transcriptome profiling-based disease classification in AS-FFPE tissue. Methods Genome-wide gene-expression profiles of AS-FFPE tissues of 83 hepatocellular carcinoma (HCC) and 47 liver cirrhosis samples were generated by using whole-genome DASL assay (Illumina), and compared with the profiles previously produced by using FC tissue sections from the same FFPE blocks. Quality of the profiles and performance of gene signature-based class prediction were systematically evaluated. Results RNA quality and assay reproducibility of AS-FFPE RNA were comparable to intermediate ~ poor quality FC-FFPE samples (R2 as high as 0.93). Gene-expression signal was detected in lower number of probes in AS FFPE samples compared to FC-FFPE samples (proportion of probes with present signal (%P-call): 10%~60% and 70%~90% in AS- and FC-FFPE profiles, respectively). Based on %P-call quality threshold of 20%, 64/88 (77%) HCC and 37/48 (77%) liver profiles were judged as having relatively good quality data with comparable inter-sample correlation. Inter-sample correlation coefficient, as a measure to detect outlier profiles due to poor RNA quality, was also lower in AS-FFPE (0.4~0.9) compared to FC-FFPE (0.6~1.0). In the genome-wide profiling analysis, previously identified molecular subclasses of HCC tumors were reproduced in 67/83 (81%) samples, which was improved to 43/48 (90%) samples when we focused on statistically confident predictions (p<0.05). A 186-gene prognostic signature in liver cirrhosis was reproduced in 32/47 (68%) samples, which was slightly improved to 11/16 (69%) when focused on statistically significant predictions. Conclusions We observed decay of genome-wide transcriptional profiles in AS-FFPE tissues in quantitative manner. However, disease classification was still possible, which suggests potential of AS-FFPE material for clinical diagnosis and prognosis.

Dataset Information

Defining glioma subtypes based on robust transcriptional patterns from 16 prior studies

Dataset's files

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets