Journal: Biotechnology Reports
Article Title: Integrative analysis of RAS signaling effectors reveals stage-dependent oncogenic patterns in colon adenocarcinoma
doi: 10.1016/j.btre.2025.e00902
Figure Lengend Snippet: Schematic overview of the analysis workflow. Publicly available gene expression datasets for RAS effectors in COAD, LUAD, LUSC, and corresponding normal tissue samples were acquired using the TCGA database with the Xena bioinformatics tool. Subsequent data transformation procedures and outlier detection were implemented to ensure data quality. These steps were followed by an analysis of single gene expression profiles. Then, Pearson’s correlation analysis was conducted to assess gene pair expression relationships across three cancer types. Colon adenocarcinoma (COAD) was selected for subsequent analyses due to exhibiting the most significant correlation shift among the cancer types. Additionally, a combinational analysis of RAS effectors in LDA visualized the separation of COAD stages based on RAS effector expression levels. Genes were ranked according to the balanced accuracy metrics derived from a logistic regression model with cross-validation, which enabled to identify potential core up- and downregulated RAS effectors that distinguish normal samples from early and late stages of COAD. Finally, key findings were validated in patient-derived COAD samples through qPCR measurements.
Article Snippet: RNA sequencing (RNA-seq) expression data for various human cancer types: COAD (normal n = 40, tumor =276), LUAD (normal n = 59, tumor n = 508), LUSC (normal n = 51, tumor n = 497) were gathered from The Cancer Genome Atlas Program (TCGA) using the University of California Santa Cruz (UCSC) Xena bioinformatic tool ( https://xenabrowser.net/ ) [ ].We specifically collected gene expression data for the RAS effectors ( ).
Techniques: Gene Expression, Transformation Assay, Expressing, Derivative Assay, Biomarker Discovery
