Hello, I want to compare diversity in two different datasets which have very large ranges of sequencing depth. The first dataset only has 3 samples, and sequencing depth ranges from 2,600-3,600 counts/sample. The second dataset has 30 samples and sequencing depth ranges from 34,000-190,000 counts/sample. For alpha diversity, I could technically rarefy the larger dataset to 2,600 and carry out the analysis from there. However, I am not sure how to approach beta diversity analysis. Should I carry out a normalization step or just leave the data as is?
I was thinking of merging the OTU table of both data sets and normalizing them with a log transformation and then proceed with beta diversity analysis. Any thoughts?