We implemented flexible taxonomic assignment of the Amplicon Sequence Variants (ASVs) via the RDP Bayesian classifier or by BLAST as well as the option to cluster ASVs with Swarm v2, a network-based clustering algorithm. As last filtering steps, we implemented optionally a tag-switching filter to remove false positives and LULU curation of ASVs/OTUs based on sequence similarity and co-occurrence rate to more abundant clusters.

We evaluate the effect of clustering and LULU curation on the quality of biodiversity inventories based on metabarcoding of ribosomal and mitochondrial barcode markers, using two eukaryotic mock communities and 42 deep-sea sediment samples.

The raw sequencing data associated to this work can be found in the /dna-sequence-raw folder, while all processing steps including config scripts and intermediate outputs can be found in the /operation folder.

The scripts for the pipeline can be downloaded from: https://gitlab.ifremer.fr/abyss-project/abyss-pipeline