I have microbial community samples taken in triplicate. The triplicates have the same metadata. I want to merge these triplicates prior to comparing them to an older dataset which has only one replicate per sample. My idea is to rarefy all the samples to the smallest library size, then merge the triplicates. Does this seem like a robust method for merging?
I have done PCA plots to look at how tightly the triplicates cluster together. They cluster very well on the OTU level, but not all of them cluster well on the ASV level. Iām interested in ASV level analyses. Would it be possible to rarefy/merge on the OTU level and then continue forward in analyses with ASVs?
Hello!
I would keep it in triplicate and do it post analysis using R. You may also want to explore this: cluster.fit.
Also, why not keep the triplicate and consider it as random effect in downstream analysis? Sorry, do not have any very constructives ideas.
2 Likes
I would also keep them separate.
1 Like