ITS + 16S sequences in the same fastq file, how to separate them before DADA2?

[lisacarraro1982 (Lisa)]

Hi,
I have vegetal ITS + bacterial 16S sequences in the same fastq library (I gave same barcode to the 2 different amplicons).

My data are PairedEndSequencesWithQuality already demultiplexed from the Sequencing Service.

I would like to separate 16S sequences from ITS sequences before running DADA2 searching the primer sequences inside the reads. 16S Primer for is TCCTACGGGAGGCAGCAGT and 16S reverse is GGACTACCAGGGTATCTAATCCTGTT.

I create a minidaset with 3 of my fastq.gz samples named 100, 101, 102.

I imported with the script qiime tools import --type ‘SampleData[PairedEndSequencesWithQuality]’ --input-path raw --input-format CasavaOneEightSingleLanePerSampleDirFmt --output-path demux-paired-end.qza

I have created a metadata file like this:

Sample Barcode

100 TCCTACGGGAGGCAGCAGT
100 GGACTACCAGGGTATCTAATCCTGTT
101 TCCTACGGGAGGCAGCAGT
101 GGACTACCAGGGTATCTAATCCTGTT
102 TCCTACGGGAGGCAGCAGT
102 GGACTACCAGGGTATCTAATCCTGTT

Can I use CUTADAPT ?

This scrips didn’t work:

qiime cutadapt demux-paired --i-seqs demux-paired-end.qza --m-forward-barcodes-file metadata.txt --m-forward-barcodes-column Barcode --output-dir demultiplexed --o-untrimmed-sequences missed.qza

Thank you very much!!

Lisa

[ebolyen (Evan Bolyen)]

Hi @lisacarraro1982,

I think I have another way to approach this.

@lisacarraro1982

I would like to separate 16S sequences from ITS sequences before running DADA2 searching the primer sequences inside the reads. 16S Primer for is TCCTACGGGAGGCAGCAGT and 16S reverse is GGACTACCAGGGTATCTAATCCTGTT.

DADA2 is actually perfectly fine with mixed amplicons, it just makes the downstream analysis more complicated, but I think there's a relatively straightforward way to handle this:

1. Run DADA2 on your mixed amplicons.

2. Use qiime feature-classifier extract-reads to pull out the ITS data for your resulting representative sequences (FeatureData[Sequence]), and then again to pull out the 16S data. Keep track of these new reads, they are your rep-seqs for anything else downstream.

3. Filter your feature-table using the newly created rep-seqs: qiime feature-table filter-features, the trick is you are going to treat your rep-seqs as metadata, so you can pass it as --m-metadata-file rep-seqs-16s.qza (or whatever you named it). If you do that once for each amplicon, you will have a pair of feature tables and representative sequences for each

[lisacarraro1982 (Lisa)]

@ebolyen

qiime feature-table filter-features

Thank you very much!