Hi there,
I’m trying to assign taxonomy to my fungal ITS1 data using a pre-trained UNITE classifier. I’m using QIIME 2 version 2025.10 and conda (miniconda). I was having issues running ITSxpress with my paired-end data (I think due to poor data quality) so have run it for single-end R1. I then denoised with DADA2. I have downloaded Git-LFS and I think successfully installed it in my terminal, but am unclear about where to get the UNITE classifier from. From a previous forum question, this was linked:
But I am not sure that the classifier there is compatible with the version of QIIME 2 I’m using. Can anyone assist? Do I need to build my own classifier?
This is my code so far:
\# FUNGUILD SINGLE-END
\# Activate QIIME
conda activate qiime2-amplicon-2025.10
\# Export existing FASTQs
qiime tools export \\
\--input-path its-demux.qza \\
\--output-path exported-demux
\# Create manifest for single-end R1
nano ITS_single_manifest.sh
#!/bin/bash
READ_DIR="exported-demux"
OUTPUT="ITS-single-manifest.tsv"
echo -e "sample-id\\tabsolute-filepath\\tdirection" > $OUTPUT
for r1 in ${READ_DIR}/\*\_R1\*.fastq.gz
do
filename=$(basename "$r1")
sample=$(echo "$filename" | cut -f1 -d"\_")
# portable realpath substitute for macOS and linux
abs_path="$(cd "$(dirname "$r1")" && pwd)/$(basename "$r1")"
echo -e "${sample}\\t${abs_path}\\tforward" >> $OUTPUT
done
\# Run it
chmod +x ITS_single_manifest.sh
./ITS_single_manifest.sh
head ITS-single-manifest.tsv
\# Import forward reads as single-end
qiime tools import \\
\--type 'SampleData\[SequencesWithQuality\]' \\
\--input-path ITS-single-manifest.tsv \\
\--input-format SingleEndFastqManifestPhred33V2 \\
\--output-path its-demux-single.qza
\# Run ITSxpress single-end
qiime itsxpress trim-single \\
\--i-per-sample-sequences its-demux-single.qza \\
\--p-region ITS1 \\
\--p-threads 8 \\
\--o-trimmed its1-trimmed.qza
\# Run DADA2 single-end
qiime dada2 denoise-single \\
\--i-demultiplexed-seqs its1-trimmed.qza \\
\--p-trim-left 0 \\
\--p-trunc-len 0 \\
\--o-table its-dada2-table.qza \\
\--o-representative-sequences its-dada2-repseqs.qza \\
\--o-denoising-stats its-dada2-stats.qza \\
\--o-base-transition-stats base-trans-stats.qza
\# Export DADA2 table
qiime tools export \\
\--input-path its-dada2-table.qza \\
\--output-path exported-table
\# Cluster to 97% OTUs
qiime vsearch cluster-features-de-novo \\
\--i-sequences its-dada2-repseqs.qza \\
\--i-table its-dada2-table.qza \\
\--p-perc-identity 0.97 \\
\--o-clustered-table its-otu-table.qza \\
\--o-clustered-sequences its-otu-repseqs.qza
\# Download Git LFS
\# Go to folder
cd \~/ (my path here)
\# Make Git LFS installation script executable
chmod +x install.sh
\# Install
./git-lfs version
