MetaWorks: A flexible, scalable bioinformatic pipeline for high-throughput multi-marker biodiversity assessments

provided in the config_ESV.yaml file but these can be customized by the user. SEQPREP was initially chosen for pairing the forward and reverse reads, because the program comes with the option to output the alignments for visual inspection, an option that most read-pairing pro- grams do not have [43]. For SEQPREP read-pairing, users can specify a Phred score quality cutoff, the minimum overlap between the forward and reverse reads, the maximum fraction of mismatches allowed in the overlap region, and the minimum fraction of matching bases in the overlap region. CUTADAPT was chosen for primer-trimming because it is fast and already widely used in the metabarcoding community for this purpose, so most users will likely already be familiar with how this program works [44]. For CUTADAPT, users need to provide a FASTA-formatted primer sequence file (adapters.fasta), they can also specify the minimum sequence length to retain after primer-trimming, a Phred quality score cutoff, the maximum error rate, minimum adapter overlap, and maximum number of ambiguous bases allowed. Fig 1. MetaWorks workflow to produce taxonomically assigned exact sequence variants. To aid reproducibility, a Conda environment is provided. Although multiple Snakemake workflows are provided in MetaWorks, here we show the main workflow that generates taxonomically assigned ESVs. Input files are shown in the first panel (green), the ESV workflow is shown in the centre panel (blue), and

outfiles are shown in the last panel (orange). The input files in white boxes are required by snakemake to run the appropriate workflow. The input files in green need to be supplied by the user. Note that only custom-trained classifiers such as for COI need to be supplied by the user whereas classifiers built-in to the RDP classifier are used automatically to process prokaryote 16S assignments, for example. The denoising step shown here includes the removal of rare clusters, sequences with putative errors, as well as chimeric sequences. The results are provided in a comma-separated value (CSV) file and shows each ESV per sample with read counts and taxonomic assignments. Abbreviations: Demultiplexed Illumina paired-end reads (R1 + R2), internal transcribed spacer (ITS) region, open reading frame sequences (ORFs). https://doi.org/10.1371/journal.pone.0274260.g001 PLOS ONE MetaWorks mult-marker bioinformatic pipeline PLOS ONE | https://doi.org/10.1371/journal.pone.0274260 September 29, 2022 4 / 11