This 90-second video shows how to use PCA to identify potential outlier samples, then verify them using other viewers such as Line Graph and Histograms.
The goal is not to remove samples automatically, but to examine why they appear different before deciding how to proceed.
See more about this dataset?
-> Does CPM Normalization "Fabricate" Batch Effects?
Learn more about PCA?
->
RNA-Seq Tutorail - PCA and Clustering