Research Paper Volume 12, Issue 14 pp 14506—14527

Figure 8. Workflow. Schematic overview of this study. Three datasets were downloaded from GEO, they are GSE18842, GSE19188, and GSE70880. A total of 287 samples were in GEO datasets. Two datasets were downloaded from TCGA, including the LUAD dataset and the LUSC dataset. Totally 216 samples were contained in those LUAD and LUSC datasets and we combined them as TCGA datasets. Datasets were divided into 3 groups based on their platforms. Affymetrix dataset and TCGA dataset were used as training sets separately then validated using the other datasets. The lncRNAs in common were used for survival analysis and functional analysis.