I am currently doing a Differential Gene Expression of RNAseq data in Lung Adenocarcinoma using TCGAbiolinks. In the data Preprocessing step TCGAanalyze_Normalisation, I am confused as to which method to select for normalization (geneLength or gcContent). What are the possible implications of both methods and how to decide which is the best suited model for the analysis?
Is there a better way to perform the DGE ?
The TCGAbiolinks reference manual provides a clear explanation of the two methods:
Normalization for RNA-Seq Numerical and graphical summaries of RNA-Seq read data. Withinlane normalization procedures to adjust for GC-content effect (or other gene-level effects) on read counts: loess robust local regression, global-scaling, and full-quantile normalization (Risso et al., 2011). Between-lane normalization procedures to adjust for distributional differences between lanes (e.g., sequencing depth): global-scaling and full-quantile normalization (Bullard et al., 2010).
Ideally, you would know if your samples spanned different lanes or not, which would indicate which you should use - "geneLength" if so, "gcContent" if not. If you're using TCGA data, that metadata is likely available somewhere.
