I have a problem that is about DADA2, I think it should have similar result, but it's not that.
I use DADA2 to run two datasets. The bacteria and fungi, but they have different effect. The result is very confusing. The bacteria-stats is for bacteria, fungi-stats is for fungi. bacteria-stats.tsv (7.0 KB) fungi-stats.csv (5.9 KB)
I want to know that why they have different clustering effect. My datasets come from the same company.
I look forward to thank you for your browse and reply.
percentage of input passed filter: mean 93% for fungi, mean 74% for bacteria
percentage of input non-chimeric: mean 80% for fungi, mean 27% for bacteria
So you are losing more bacteria during quality filtering, which could just be an unlucky run for bacteria. However, you are also losing many more reads to chimera filtering for bacteria, which could be worth investigating!
What region(s) did you sequence for each Kingdom? What trimming and truncation settings did you use when running DADA2 on each amplicon?
Hi, thanks for your reply, firstly, I sequence my data for full-length, and they are third-generation sequencing. Secondly, before I use DADA2, my data have been trimed and truncated, so I don't use trimming and truncation settings.
eg : --p-trim-left 0 --p-trunc-len 0
Did you analyse the quality of the reads F and R? I had same issue and it was due to the low quality of the R reads... My F reads had 350 nt with good quality average so I decided to truncate them at that length and just work with F reads... It change from around 20% of non-chimeric to around 85%...
Really? The full length genes are longer than would allow pairing on the Illumina platform, so it may be worth double checking this...
It's recommended to run all trimming and preprocessing within Qiime2 so that you have more control of these settings and can track all changes made within the provenance platform. But that's up to you, and probably would not explain differences in chimera levels.