ITS quality control and denoising

[ari_sh70 (Armin)]

Hello everyone,

I have 48 paired-end read ITS samples. I used the quality control and denoising like the way that I have used for 16S data but the result in ITS wasn't good like 16S. Did I choose the wrong method for quality control? I would appreciate it if anyone can help me, please.

The script that I have used for quality control:

qiime dada2 denoise-paired
--i-demultiplexed-seqs demux-paired-end.qza
--p-trim-left-f 0
--p-trim-left-r 2
--p-trunc-len-f 285
--p-trunc-len-r 285
--o-table table.qza
--o-representative-sequences rep_seqs.qza
--o-denoising-stats denoising_stats.qza

The interactive plot:

The output:

One more question: What if I join the forward and reverse reads and then I do the quality control, wouldn't be better?

Thanks,

Armin

[Nicholas_Bokulich]

The issue is that the quality is not great, and you are not truncating enough. So you lose a lot of reads at the filtering stage and at the merging step.

The merging step probably can’t be helped… ITS reads have difficulty merging because the length is very variable and so some will merge while others will fail. In the end it may be wisest to just use the forward reads.

You should truncate a little more — maybe to 240? To get more reads passing filter.

There is heaps of guidance on dada2 truncation troubleshooting on this forum, and ITS merging issues, and other ITS processing issues. I recommend using the forum search tool to find and read through these past discussions to see what others have done!

[ari_sh70 (Armin)]

Thank you @Nicholas_Bokulich for your reply and time.

The merging step probably can’t be helped… ITS reads have difficulty merging because the length is very variable and so some will merge while others will fail. In the end it may be wisest to just use the forward reads

Yes exactly but preferably I would like to use both.

You should truncate a little more — maybe to 240? To get more reads passing filter.

Sorry, I didn’t understand this. I’ve used 285 for trunc. What do you mean by 240?

There is heaps of guidance on dada2 truncation troubleshooting on this forum, and ITS merging issues, and other ITS processing issues. I recommend using the forum search tool to find and read through these past discussions to see what others have done!

For sure, I did and will continue to read more and more!

Again, thank you for your time and answers.

Best,
Armin

[Nicholas_Bokulich]

Sorry, I didn’t understand this. I’ve used 285 for trunc. What do you mean by 240?

Nicholas_Bokulich:

try truncating at 240 and more of your reads may pass the filtering step, since you will be excluding more of the noisy 3’ ends of the sequences.