Normalized or Absolute counts for comparitive study?

User Questions
1

Dear Megan Community member,

I have control and treated RNASeq samples. I want to use them for metatranscriptomics analysis.
I have generated .daa file using SAMSA 2 package.
I have meganized the .daa file using MEGAN 6 software.
I would like compare the above 2 datasets.
In the compare panel, should I use absolute reads or normalized reads? Which would be a better analysis?
The normalized counts is reducing to 4 lakhs whereas the absolute reads is upto 32 lakhs

Kindly help me through this.
Thanks in advance!

2

The normalized comparison normalizes to the smallest given count. If you don’t normalize, then the “species (or taxa) richness” of your samples with be biased, with the smaller samples appearing less species-rich due to the fact that more low abundance taxa are missing.

3

Dear Daniel,
Thank you for the clarification, I will use the normalized data.