MARCO-BOLO Data Analysis Challenge

Short description: In September 2024, MARCO-BOLO launched the Data Analysis Challenge to survey the global environmental DNA (eDNA) community about their use and development of bioinformatics pipelines to process eDNA samples, as well as invite them to contribute their time and methodology towards an extensive comparison of these pipelines. Registered participants were invited to download any of the four metabarcoding datasets provided, analyse these with their pipeline, and provide back the output ASV/OTU and taxonomy tables for comparison. In total, 13 participants submitted a total of 15 output datasets for the 18S protist observatory dataset, and 25 participants submitted a total of 101 output dataset for the 12S, 16S and COI fish aquarium datasets. The resulting tables were harmonised to be analysed conjointly, and bioinformatics parameters were extracted from the different results files we received. The current data submission consists of the harmonised tables of participant outputs, the extracted bioinformatics metadata, and some supplementary files, all detailed below. Metadata: dataset_description.json describes all the datasets available in this repository, as well as each field within each dataset. Original data: 18S protist metabarcoding from the ASTAN observatory, published by Caracciolo et al. (2022) Metabarcoding raw reads available on the EuropeanNucleotide Archive (ENA): PRJEB48571 Microscopic counts file available on Zenodo (Rigaut-Jalabert et al. 2021): https://doi.org/10.5281/zenodo.5033180. These were restructured to fit the structure of the challenge datasets, and are provided here as microscopy_harmonised.csv. 12S, 16S, COI fish metabarcoding from aquarium samples. The metabarcoding raw reads were communicated to participants directly. These are currently unpublished, and owned by the A-Fish-DNA-Scan and ME-BARCODE group at the University of Minho. The lists of fish, shark and ray species inhabiting the Lisbon aquarium at the time of sampling (“residents”), and species used in feed are given here (aquarium_12s16sCOI_residents.csv & aquarium_12s16sCOI_feed.csv) Sampling area: The 18S protist metabarcoding samples were collected at the SOMLIT-Astan sampling station (http://marineregions.org/mrgid/5372) in the Western English Channel, France (60 m depth, 48°46’18” N–3°58’6” W). The 12S, 16S, COI fish metabarcoding samples were collected from the Lisbon Aquarium, Portugal (38°45’48.671” N-9°5’37.377”). Instructions: Instructions to participants were communicated and archived on the public repository https://github.com/marco-bolo/wp2-wp5-workshop. Participant output data: The resulting output tables from each participant, harmonised and merged by target group (protists or fish), are available in challengedata_18s.csv and challengedata_12s16sCOI.csv. Participant bioinformatics workflows: Key bioinformatic steps and parameters were extracted for each of the submissions included in the challenge data. metadata_18s.csv: bioinformatics metadata accompanying the challengedata_18s.csv metadata_12s16sCOI.csv: bioinformatics metadata accompanying the challengedata_12s16sCOI.csv The selected fields describing the bioinformatics steps and parameters were taken from the FAIR eDNA checklist (Takahashi et al. 2025). Output data analysis: All scripts to compare the participant outputs, analyse patterns and generate figures are being published to the public repository https://github.com/marco-bolo/data-analysis-challenge. Status: Dataset is marked incomplete — analysis is ongoing but data challenge participation is complete (September 2024 – November 2025). Currently, participants are indicated with an anonymized identifier (submitterID). Names will be released upon agreement of all participants in a next version.