Hi,
I am analyzing 16S rRNA gene sequencing data from a low biomass sample using QIIME2 and DADA2 for preprocessing.
Our primary research question is: Is a specific bacterial taxon present in a particular sample?
One challenge we are facing is determining whether an ASV observed at low relative abundance is truly present in the sample or merely a contaminant or artifact (e.g., from PCR/sequencing error). Unfortunately, we do not have any positive or negative controls in this dataset to help identify background noise or contaminants.
We are considering filtering out low abundance taxa, using relative abundance thresholds of 0.01%, 0.1%, or 1%—based on what has been done in previous studies.
My specific questions are:
Any insights or recommendations would be greatly appreciated.
Regarding "artifact (e.g., from PCR/sequencing error)", here is a paper for you reference. Article Minimizing spurious features in 16S rRNA gene amplicon sequencing
Regarding "merely a contaminant", it is better to have negative and positive controls especially for low-biomass samples.
This question is mixing two different things.
Data analysis in DADA2 and further down-stream analysis.
Filtering low abundance is not the part in DADA2, it is more or less visualization part.
And for the specific question in point 1-4, these are subjective questions which can't be and shouldn't be generalized. Filtering or not would depend on the specific datasets and the questions around the project. There can be several ways these question can practically be handled, but the aim always precede the practical workflow.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers