search

benjjneb / decontam

Simple statistical identification and removal of contaminants in marker-gene and metagenomics sequencing data
https://benjjneb.github.io/decontam/
148 stars 25 forks source link

Decontam w/ just one control? #38

Closed andrebolerbarros closed 5 years ago

andrebolerbarros commented 5 years ago

I'm now starting to use the package in the analysis of 16s amplicon sequencing. I've read through the paper and the tutorial and it sounds very useful for me indeed!

However, the data that I've now to analyse only contains one control sample (negative) per set of samples. Although I can still work with the frequency method; however, I was wondering if there is an alternative to the prevalence method in this conditions.

These are indeed far from ideal, I know but, I'm just trying to see if I can still do some contaminant filtering in these samples.

Thanks!

benjjneb commented 5 years ago

No, the prevalence method is not useful with just a single negative control, and in general I would recommend caution in removing features based on one negative control because of the possibility of cross-contamination putting real ASVs into negative controls. The way I would use a single negative controls would be just as a post-hoc check on any conclusions of my main analysis. E.g. if I find ASV22 to be highly predictive of my condition of interest, I would check if its also in that control and at least mention it in the text if it was.

The frequency method should work just fine for you though, and is effective all on its own.

andrebolerbarros commented 5 years ago

Thanks @benjjneb for the swift answer!