Data and scripts for "Full-length COI barcodes improve eDNA metabarcoding data denoising relative to mini-barcodes"

This repository contains the processed data and the R and Python scripts used for the statistical analyses and figures of "Full-length COI barcodes improve eDNA metabarcoding data denoising relative to mini-barcodes" (Molecular Ecology Resources). Soil invertebrate communities from 34 samples were sequenced for three COI amplicons: 313 bp (Illumina), 660 bp and 1,256 bp (PacBio). Workflow: Raw reads (NCBI SRA, BioProject PRJNA743174) were processed with PipeCraft2 (Anslan et al. 2017) to obtain ASV tables, ASV sequences and SINTAX taxonomy (data/1.ASV_tables_and_taxonomy.RData). Animalia ASVs were clustered into OTUs and filtered with metaMATE to remove non-authentic sequences such as NUMTs and other artefacts (data/2.ASVs_Animalia_to_metaMATE.RData, data/3.Final_OTU_tables_and_taxonomy.RData). The scripts compare the three amplicon sets in terms of metaMATE classification, shared OTUs between amplicon sets, taxonomic assignment confidence, OTU richness and community composition (NMDS, PERMANOVA, Procrustes), and stop-codon detection in relation to sequence length. The README describes all data files, the order in which to run the scripts, the figure and table each script produces, and the software dependencies.