Biomarker Benchmark - GSE20189

Version 6 2016-03-17, 22:19

Version 5 2016-03-17, 21:25

Version 4 2016-02-23, 23:35

Version 3 2016-02-04, 22:17

Version 2 2016-02-02, 22:08

Version 1 2016-02-02, 22:05

dataset

posted on 2016-03-17, 22:19 authored by Anna GuyerAnna Guyer, Stephen PiccoloStephen Piccolo

[NOTICE: This data set has been deprecated. Please see our new version of the data (and additional data sets) here: https://osf.io/mhk93 ]

"Affordable early screening in subjects with high risk of lung cancer has great potential to improve survival from this deadly disease. We measured gene expression from lung tissue and peripheral whole blood (PWB) from adenocarcinoma cases and controls to identify dysregulated lung cancer genes that could be tested in blood to improve identification of at-risk patients in the future. Genome-wide mRNA expression analysis was conducted in 153 subjects (73 adenocarcinoma cases, 80 controls) from the Environment And Genetics in Lung cancer Etiology (EAGLE) study using PWB and paired snap-frozen tumor and non-involved lung tissue samples. Analyses were conducted using unpaired t-tests, linear mixed effects and ANOVA models. The area under the receiver operating characteristic curve (AUC) was computed to assess the predictive accuracy of the identified biomarkers. We identified 50 dysregulated genes in stage I adenocarcinoma versus control PWB samples (False Discovery Rate ≤0.1, fold change ≥1.5 or ≤0.66). Among them, eight (TGFBR3, RUNX3, TRGC2, TRGV9, TARP, ACP1, VCAN, and TSTA3) differentiated paired tumor versus non-involved lung tissue samples in stage I cases, suggesting a similar pattern of lung cancer-related changes in PWB and lung tissue. These results were confirmed in two independent gene expression analyses in a blood-based case-control study (n=212) and a tumor-non tumor paired tissue study (n=54). The eight genes discriminated patients with lung cancer from healthy controls with high accuracy (AUC=0.81, 95% CI=0.74-0.87). Our finding suggests the use of gene expression from PWB for the identification of early detection markers of lung cancer in the future."

http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE20189

We have included gene-expression data, the outcome (class) being predicted, and any clinical covariates. When gene-expression data were processed in multiple batches, we have provided batch information. Each data set is organized into a file set, where each contains all pertinent files for an individual dataset. The gene expression files have been normalized using both the SCAN and UPC methods using the SCAN.UPC package in Bioconductor (https://www.bioconductor.org/packages/release/bioc/html/SCAN.UPC.html). We summarized the data at the gene level using the BrainArray resource (http://brainarray.mbni.med.umich.edu/Brainarray/Database/CustomCDF/20.0.0/ensg.asp). We used Ensembl identifiers. The class, clinical, and batch data were hand curated to ensure consistency ("tidy data" formatting). In addition, the data files have been formatted to be imported easily into the ML-Flex machine learning package (http://mlflex.sourceforge.net/).