Few sequences left after denoised

User Support
1

Dear all,

Can anyone advise me about this issue? After running the command below
qiime dada2 denoise-single --i-demultiplexed-seqs qm/demux.qza --p-n-threads 64 --p-trim-left 0 --p-trunc-len 0 --o-representative-sequences qm/rep-seqs.qza --o-table qm/table.qza --o-denoising-stats qm/dada2-stats.qza --verbose

I got only few denoised sequences of some samples. Would it be possible to adjust the command to get more sequences?

image
image1455×767 35.8 KB

Thanks!

2 Likes
2

Hi @shengxinfenxi,
Can you upload a picture of the interactive quality plot in your demux.qzv? I think that more conservative truncation values might fix this, but I will need to see the interactive quality plot to confirm.

3

Thanks@cherman2, the picture of the interactive quality plot and demux.qzv has been uploaded.

I can share the others as well if that would help, just let me know. Thanks again for your help!

image

demux.qzv (310.4 KB)

4

I noticed that you have really long sequences!

Can you tell me what sequencing technique you used to generate them?

5

These sequences were ccs generated by Pacbio sequel II. Before importing qiime2, I had removed the barcode and primer sequence, and filtered out reads smaller than 1200 or larger than 1800.
I have also tried using this command to denoise and obtained similar results as before.
qiime dada2 denoise-ccs --i-demultiplexed-seqs demux.qza --p-n-threads 64 --o-representative-sequences rep-seqs.qza --o-table table.qza --o-denoising-stats dada2-stats.qza --verbose --p-front ********* --p-adapter ********* --p-min-len 1200 --p-max-len 1800

6
7

Hi @shengxinfenxi,
Can you post the stats file you got from the qiime dada2 denoise-ccs command?

1 Like
8
9

The reads after denoised have changed, but the fundamental issue has not been resolved.

By the way, how can I get the data, which have been removed adapaters, primers and filter the reads beyond 1200-1800, while using the deonised ccs command.

image
image1103×788 42.1 KB

demux.qzv (502.7 KB)

10
11

Hi @shengxinfenxi,
These seem like reasonable results. Unfortunately, Some samples have a high percentage of chimera, like that top sample.

I am not sure what you mean here. Can you elaborate?

1 Like
12
13

The first sample seems to have been removed a lot of reads in the deonised step. I have another batch of samples, all of which show that a lot of reads have been discarded in the deonised step (denoise-ccs). Is there any command or parameter that can improve this situation?

image
image1848×480 28 KB

The denoise-ccs command will remove the primers and adapters of reads and filter reads based on their max and min lengths. I want to obtain the clean data after removing primers and adapters, and filtering some reads based on length parameters.Clean data is very important to me. Could you tell me how to get these clean data.Thanks!
demux.qzv (370.3 KB)

14
15

Hi @shengxinfenxi

Yes I agree. One of your samples has a lot of reads removed due to denoising. However, sometimes samples just have bad quality and there is nothing that can really be done. My rule of thumb is if all of my samples lost 50% or more to denoising, then I am worried something went wrong.

If you want to mess around parameters there are a couple of parameters that you could test out.

  1. You could try trimming at a higher value, right now it seems like you are not trimming at all. Trimming before your interactive quality plot starts dropping.
  1. You could increase max expected error parameter: --p-max-ee. Just a warning though as you increase the max-expected-error threshold you will get more errors in your sequences.

The rep-seqs that you get out of denoised-ccs should be your clean reads.

Hope that helps!
:turtle:

1 Like
16
17

I wanted to take a moment to express my heartfelt gratitude for sharing your valuable experiences and insights on the forum. Your contributions have been incredibly helpful to me.

The clean data I want is the data after removing primers and adapters for each sample, but before denoising.

3 Likes
18

Hi @shengxinfenxi,

You are able to get the data you're looking for by running the following command:
qiime cutadapt trim-single

3 Likes
19

Thanks for your advise. The cleandata I want is generated through the intermediate process of the following command. This command can remove adapters, primers, and filter reads based on minimum and maximum lengths (1200-1800). Then it began to denoise the sequence.
“qiime dada2 denoise-ccs --i-demultiplexed-seqs demux.qza --p-n-threads 64 --o-representative-sequences rep-seqs.qza --o-table table.qza --o-denoising-stats dada2-stats.qza --verbose --p-front ********* --p-adapter ********* --p-min-len 1200 --p-max-len 1800”.
I also checked the command you mentioned.
"qiime cutadapt trim-single"

image
image487×853 23.7 KB

This command can remove adapters, primers, and filter reads based on the minimum length, but it seems that it cannot filter reads based on the maximum length of the library. Perhaps I did not use this command correctly. Could you give me additional advice. Thanks again.

20

Hi @shengxinfenxi,
I am kinda under the impression that we do not have the functionality you want.
Can you elaborate why you want to through out reads that are over a specific length?

1 Like