too many zeros in 16S rRNA amplicon sequencing data
Hi all,
I have some 16S rRNA sequencing data from human tissue, which was performed on NovaSeq 6000 platform. I detected ~40000 ASVs in total after removing singletons. However, most ASVs were only present in a few samples. In other words, ~95% ASVs are Zero in one specific sample. Those huge numbers of Zero significantly prevent me from downstream analysis.
Whether removing low frequent ASVs is also a controversial topic. Could anyone give me some advice? Thanks!
• 1,816 views
•
link
written
by
zhangdengwei
0 answers
No answers yet.
Log in to answer this question.
More posts like this
-
rarefaction curve metagenomic reads show species saturation around 100 000 reads
written by rDNA •
-
Controlling for Baseline Taxa Relative Abundance [Longitudinal 16S rRNA Sequencing Data]
written by bjeremy •
-
Tool: Introducing NovaDemux, a Sequence Demultiplexer that Increases Yield and Salvages Failed Runs
written by Brian Bushnell
-
should remove unassigned genera in 16S rRNA sequencing analysis
written by zhangdengwei
-
Question about 16S rRNA reads we got from a sequencing company
written by KQUB •
-
Blog: ANTI-mosquito and Healthcare —— How Next Generation Sequencing Will Support Research of Vecto…
written by Novogene •
-
Feature Selection in Large Dataset with High Multicollinearity
written by KernelGhost •
-
Combining own and public samples from different sequencing strategies to perform GWAS
written by antmantras •
-
DE analysis with limma (What to include in the model?)
written by ngs101.com
-
Microbiome decontamination from kitome
written by robert.kwapich •
You have too many ASV (40k) and I think this is a problem related to the denoising (DADA2) of NovaSeq 6000 sequencing data. Maybe this post could be helpful: link
Exactly. I agree that this is due to the NovaSeq which is not very compatible with qimme2 DADA2 pipeline. I finally surrendered to other approaches.