Detalle de: Identification of reference genes in lung cancer from RNA-seq data

Introduction: Gene expression analysis need to use reference genes (RGs) to normalize the measurements. Ideally, a RG should be expressed in the same degree in all the studied tissues, that is why housekeeping genes usually are used. But it has been demonstrated that they are less invariant than what was thought. Objectives: To find a set of RGs in lung cancer. Methods: RNA-seq reads of healthy and tumoral samples from EGA (17 patients with small-cell lung cancer [SCLC]) and NCBI (50 patients with adenocarcinoma lung cancer) repositories were downloaded. Reads were processed with a novel algorithm developed by us based on the coefficient of variation (CV) to select the genes with a CV lower than 15% or 20%, and whose number of reads per transcript was higher than 10,000. Hence, the most stable RGs having the highest expression level were selected. Results: Using 10,000 reads and CV < 15% as filter, 7 genes in adenocarcinoma and 3 in SCLC were considered RGs. Using CV < 20%, 459 RGs were found in adenocarcinoma and 21 in SCLC. If we mix all samples then we find 27 genes with CV < 20% and only 1 with CV < 15%. Those RGs contained some already known RGs for other cancers, like the ZNF26 and ZNF556 (zinc finger proteins). Conclusions: The set of reference genes has been extended in two lung cancers, which makes more detailed expression analysis possible.