OTU read counts

[kristin oosthuizen]

Hi,

I have now created my OTU table and I just want to be sure I understand it right. The relative abundance is measured in read counts right? I am literally taking it one step at a time with my data analysis and I was wondering is there a script or command to normalize the read counts (RPKM)? I want to perform diversity and stats, but was wondering if these counts are normalized at some point further down the pipeline?

Thanks!
Kristin

[justink]

Sounds good. This is 16S amplicon sequencing, yes? If so, I wouldn't normalize like RPKM—16S genes tend to be relatively similar in length (compared to the variety of all transcripts in shotgun RNAseq where RPKM is often used).

I'd say plow ahead with the unnormalized read counts. Many of the measure to compare samples (e.g. unifrac) operate on unnormalized (but rarefied!) sequence counts.

[kristin oosthuizen]

Hi,

Thank you for your response. No actucally I did ITS amplicon sequencing. I just wondered if I should normalize the read counts, or if I should just ago ahead and run core diversity analyses?

Thanks!
Kristin

[justink]

Do the core diversity analyses! That'll handle rarefaction / normalization.