Appropriate use of DESeq2-VST for sampling design

I wanted to use the Variance Stabilizing Transformation function through the DESeq2 package in R, but I am wondering what is the most appropriate way to go about this with my sampling design. I have triplicate environmental samples from 8 sites, but I also ran controls at each step to identify contaminants. I have a field blank, extraction controls and sequencing controls along with all of my samples that were sequenced via Illumina MiSeq. I am looking for guidance on if I should manually remove the identified contaminant sequences prior to using DESeq2-VST transformation, or if there is another way to go about this that is more appropriate given how the transformation is calculated?
I am doing this transformation prior to calculating weighted beta diversity measures and visualizing through Principal Coordinate Analysis ordinations.

1 Like