Entering edit mode
I have been working on a large scale RNA-seq data analysis that are generated over 5 years. We don't have control genes and our phenotypes are continuous variables. We are trying to find the unwanted variables without depending on any phenotypes or control samples/genes. We have tried clustering/PCA. The clusters were not clear to us. Welcome any suggestions! Thank you very much!
Tongjun
Is it possible for RUVSeq to use 6 control samples to estimate the unwanted variables for more than 1K samples? Thanks!
Tongjun