cutadapt cannot remove primer in the middle of reads

General Discussion
,
1

Hello everyone,

I apply qiime2 to my 16S rRNA sequencing data, which is an amplification of the V4 region with a primer pair of "GTGCCAGCMGCCGCGGTAA" and "GGACTACHVGGGTWTCTAAT". Whereas I use the cutadapt plugin to remove the primer with different commands below, some primers still remain in the middle of sequences.

The commands I have tested are as follows:

qiime cutadapt trim-paired --p-cores 20 --i-demultiplexed-sequences paired-end-demux.qza --p-front-f GTGCCAGCMGCCGCGGTAA --p-adapter-f TTACCGCGGCKGCTGGCAC --p-front-r GGACTACHVGGGTWTCTAAT --p-adapter-r ATTAGAWACCCBDGTAGTCC --o-trimmed-sequences paired-end-demux_de_primer.qza

qiime cutadapt trim-paired --p-cores 30 --i-demultiplexed-sequences paired-end-demux.qza --p-front-f GTGCCAGCMGCCGCGGTAA --p-adapter-f TTACCGCGGCKGCTGGCAC --p-front-r GGACTACHVGGGTWTCTAAT --p-adapter-r ATTAGAWACCCBDGTAGTCC --p-anywhere-f GTGCCAGCMGCCGCGGTAA --p-anywhere-r GGACTACHVGGGTWTCTAAT --o-trimmed-sequences paired-end-demux_de_primer.qza

qiime cutadapt trim-paired --p-cores 80 --i-demultiplexed-sequences paired-end-demux.qza --p-front-f GTGCCAGCMGCCGCGGTAA --p-front-r GGACTACHVGGGTWTCTAAT --p-match-read-wildcards --p-match-adapter-wildcards --o-trimmed-sequences paired-end-demux_de_primer.qza

The result is subjected to DADA2.

qiime dada2 denoise-paired --p-n-threads 20 --i-demultiplexed-seqs paired-end-demux_de_primer.qza --p-trunc-len-f 0 --p-trunc-len-r 0 --o-table dada2_table.qza --o-representative-sequences dada2_rep_set.qza --o-denoising-stats dada2_stats.qza

denoising statistics

qiime metadata tabulate --m-input-file dada2_stats.qza --o-visualization dada2_stats.qzv
qiime feature-table tabulate-seqs --i-data dada2_rep_set.qza --o-visualization rep-seqs.qzv

The length of merged sequences is supposed to be ~250 bp, but some merged sequences are longer than 300 bp. I found that the primer "GTGCCAGCMGCCGCGGTAA" still remained in some sequences, like the case below.

>b4559da283e0265b85157cd7532d5f37
TTAGAAACCCTTGTAGTCCATTGGCGTACG***GTGCCAGCCGCCGCGGTAA***TACGTAGAAGACTAGTGTTAATCATCTTTATTAGGTTTAAAGGGTACCTAGACGGTAAATTAAACTCTAAATGAGTACTTGTTTACTAGAGTTTTATGTAAGGAGGAAGAATTTCTGGAGTAGTGATTTAATATGAATAATCTCAGAGAGACTGGTAACGGCGAAGGCATCCTTCTATGTAAAAACTGACGTTGAGGGACGAAGGC

Can anyone provide suggestions on how to address this issue? Thanks.

2

Good afternoon,

Cutadapt has lots of features, and choosing the ones that are just right for your reads can be hard.

One option to try is --p-times, which is set to 1 by default, but can be increased for cases when the primer appears a second (or third??) time inside the read.

I try to work with the sequencing core to select cutadapt settings. They will know what to expect in the read and may already have a recommended way of removing unwanted primers.

1 Like
3

Hi Colin,

Thanks for the suggestion. I address this issue with the following command:

qiime cutadapt trim-paired --p-cores 80 --i-demultiplexed-sequences paired-end-demux.qza --p-anywhere-f GTGCCAGCMGCCGCGGTAA --p-anywhere-r GGACTACHVGGGTWTCTAAT --p-match-read-wildcards --p-match-adapter-wildcards --p-times 2 --p-minimum-length 180 --o-trimmed-sequences paired-end-demux_de_primer.qza

Besides, I am curious about how this happened. Are they chimera that DADA2 is unable to recognize?

4

Maybe a primer dimer? I don't see primers twice, but that option is there because it can happen.

Does adding --p-times 2 seem to help or do you get similar issues?