Removing Zero Counts from results(dds, contrast=c("condition", "x", "y")) vs removing Zero Counts from Count Table
1
0
Entering edit mode
@tborchardt2-12444
Last seen 7.1 years ago

I am using DESeq2 to investigate changing microbial taxa after incubation experiments. I am comparing my Preaddition community to 5 post-incubation communities (different nutrient additions & Control) and looking for taxa that significantly change. I have approximately 49,000 OTUs (Operational Taxonomic Units) in the original count table which includes all of the incubations, and I am individually investigating the different incubations. To remove some of these extra OTUs I run:

> microcosmcountTable4 <- microcosmcountTable4[rowSums(microcosmcountTable4)>0,]

I go from 49,000 OTUs to 8500 OTUs meaning about 40,000 OTUs have zero counts in that particular incubation table. If I do not remove those 40k OTUs I end up getting quite different results.

So, my question stems from the command where I am investigating particular nutrient additions from Microcosm 4 count table. I run the command:

> iron4 <- results(dds4, contrast=c("condition","iron","preaddition"))

This investigates just the iron vs. the preaddition (I believe) and does not include the other nutrient treatments. Should I attempt to remove the extra zero count OTUs that likely come about from subsampling within microcosmcountTable4 since I am looking just at the iron vs. Preaddition? Is filtering at the beginning with the other nutrient additions enough? I am am unsure if there is a way to filter at this point.

Thank you!

deseq2 • 1.4k views
ADD COMMENT
1
Entering edit mode
@mikelove
Last seen 48 minutes ago
United States

Read the vignette section on "independent filtering". The short answer is that you don't have to remove the extra small counts when you are comparing two groups, because this is performed automatically when you build a results table. See the vignette and ?results for details.

ADD COMMENT

Login before adding your answer.

Traffic: 728 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6