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

Workflow The pipeline begins with demultiplexed Illumina paired-end reads as this is the format most often provided by sequencing centres to their clients. Several workflows are available as Snake- make pipelines such as taxonomic assignment of ESVs (Fig 1), clustering of ESVs into OTUs, or for processing single reads. For each of these workflows described below, parameter settings for each bioinformatic step can be customized in the config.yaml file. The user also needs to provide a file of primer sequences so we provide a template for the adapters.fasta file as well as a small set of raw Illumina sequences for the COI amplicon that can be used to test the installa- tion. The online documentation provides a tutorial example using the provided COI test data. The tutorial also walks users through the steps necessary to set up their environment to run the pipeline for the first time, assuming the user has never worked with Conda or Snakemake before. Exact sequence variants The ESV workflow will run the pipeline shown in Fig 1. In the config_ESV.yaml file, users indicate the path to the directory that contains the demultiplexed Illumina paired-end reads, specify the unique part of filenames to distinguish between samples and reads, and specify the name of the directory that will contain the outfiles. Default settings for each program are Table 1. RDP-trained reference sets that can be used with MetaWorks. Marker Target taxa Classifier availability Number of included sequences Number of included taxa at all ranks (species) Source data

COI Eukaryotes https://github.com/terrimporter/CO1Classifier 1,221,528 154,351 (114,687) BOLD [34], INSDC [35] rbcL Diatoms https://github.com/terrimporter/ rbcLdiatomClassifier 3,504 1,432 (1,023) Diat.barcode [36] rbcL Land plants https://github.com/terrimporter/rbcL_ landPlant_Classifier 148,258 61,398 (50,778) INSDC [35] rbcL Eukaryotes https://github.com/terrimporter/rbcLClassifier 164,454 65,742 (53,344) INSDC [35] 12S Fish https://github.com/terrimporter/ 12SfishClassifier 2,853 4,751 (2,833) MitoFish [37] 12S Vertebrates https://github.com/terrimporter/ 12SvertebrateClassifier== 10,654 15,007 (9,564) INSDC [35] and MitoFish [37] SSU (18S) Diatoms https://github.com/terrimporter/ SSUdiatomClassifier 2,962 1,198 (828) Diat.barcode [36] SSU (16S) Vertebrates https://github.com/terrimporter/ 16SvertebrateClassifier 72,195 21,282 (15,155) INSDC [35] SSU (18S) Eukaryotes https://github.com/terrimporter/18SClassifier 42,301 7,504 (5,440 genera) SILVA [38] SSU (16S) Prokaryotes Built-in to the RDP classifier� 13,212 3,247 (2,506 genera) RDP [5] ITS Fungi (Warcup) Built-in to the RDP classifier 17,878 10,621 (8,551) Deshpande et al., 2016 [39] ITS Fungi (UNITE 2014) Built-in to the RDP classifier 145,019 23,222 (20,337) Abarenkov et al., 2010 [40] ITS Fungi (UNITE 2021) https://github.com/terrimporter/UNITE_ ITSClassifier 1,393,203 376,167 (352,588) UNITE [40] ITS Plants https://github.com/terrimporter/PLANiTS_ ITSClassifier 104,387 72,632 (61,693) PLANiTS [41] and UNITE [40]

LSU Fungi Built-in to the RDP classifier 11,442 2,633 (1,895) Liu et al., 2012 [42] https://doi.org/10.1371/journal.pone.0274260.t001 PLOS ONE MetaWorks mult-marker bioinformatic pipeline PLOS ONE | https://doi.org/10.1371/journal.pone.0274260 September 29, 2022 3 / 11