Primers removal and if DADA2 is able to remove them

Hey guys,

My questions are sort of a extension of this other topic Question about primers/cutadapt (reverse-complement).

I am currently with a dataset of paired-end reads (forward reads have ~300bp and reverse reads ~300bp). After looking into the forward sequences with grep, it didn't return any primer (even searching into the reverse one). Here follow the primer pairs:

16s rRNA primers MSfw: 5’-CCGGGTATCTAATCCGGTTC-3’
#and MSrev: 5’- CTCCCAGGGTAGAGGTGAAA-3’.

The problem is: when searching for the reverse primer it returned the sequence in some samples, but ONLY with the reverse complement of the reverse primer (generated through https://reverse-complement.com/). When searching with the reverse primer itself it didn't return anything.

Here follows what I searched for:

grep -c "TTTCACCTCTACCCTGGGAG" *.fastq #reverse complement

Output:
SRR1_2.fastq:194
SRR2_2.fastq:4
SRR3_2.fastq:26
SRR4_2.fastq:31
SRR5_2.fastq:53
SRR6_2.fastq:58
SRR7_2.fastq:33
SRR8_2.fastq:1137
SRR9_2.fastq:361

But some samples returned 0.

And I also had already analysed this dataset, but now I am wondering if I shoud reanalyse it again but running cutadapt before the downstream analysis in order to prevent primers to interfere with the analysis and if so, how should I perform this cutadapt. With the paired-end option (and the reverse complement of the reverse primer), even if the forward primer was not detected?

Or do you think DADA2 was capable of removing them (since there was this parameter --p-trunc-len-r 241)? Is there a way to verify this?

Here are the parameters used for DADA2:

qiime dada2 denoise-paired
--i-demultiplexed-seqs paired-end-demux.qza
--p-trim-left-f 6
--p-trim-left-r 6
--p-trunc-len-f 273
--p-trunc-len-r 241
--o-table table.qza
--o-representative-sequences rep-seqs.qza
--p-min-fold-parent-over-abundance 16
--o-denoising-stats denoising-stats.qza
--verbose

EDIT: I tried to perform cutadapt just to see how the samples would retrieve and the output was almost empty (regarding the sequence counts); so I am guessing I will stay with the previous result anyway..

Thank you again in advance.

Best regards.

Hi @Liviacmg,
This post regarding cut adapt vs denoising should be helpful to you in your decisions with your analysis.
Additionally, this tutorial goes over how to use cutadapt if you chose to give that another try.
Based on this documentation I think you are going in the right direction with your thoughts on truncation!
--Hannah

1 Like