Evaluation of methods for RNA-Seq analysis for uncovering key components of estrogen receptor-alpha signaling pathway in breast cancer

Breast cancer is the most common female cancer worldwide. Higher estrogen receptor (ER) expression is often associated with poor prognosis in ER positive breast cancer, however the exact mechanism is unknown. RNA-Seq data of three different experiments of ER knockdown (siE1, siE2, siE3) was used by researchers previously to identify TNFAIP1/BACURD2 as the mediator of ER induced increase in cell migration typical of breast cancer. We herein present a more comprehensive analysis of the data using DESEq2, along with comparison of results using a non-parametric approach, SAM-Seq, in order to cover the low sample size used in the study, and compared the results. We have found that, SAM-Seq uncovers more significant genes and is as robust as DESeq2 in discovering genes deemed significant by DESeq2. Excitingly, our approach was able to uncover three most significantly DE genes among the three independent experiments, namely, UHMK1, ACLY and CLIC4. The fact that they are involved in cancer regulation, metabolism and cell signaling, but so far has barely been studied, serves as additional exciting avenues of targeting ER pathway in breast cancer. Lastly, neither of our DESeq2 nor SAM-Seq analysis results showed consistent downregulation of TNPAIP1 upon ERR-alpha knockout across three independent experiments. Since our analysis approaches are both more conservative and robust in situations of model assumption violations, we came to the conclusion that further experiments are needed to ascertain the involvement of TNFAIP1 in the ER signaling pathway in breast cancer.