Table 1 Published reference databases commonly used for taxonomic assignment in COI eukaryotic metabarcoding studies.

Reference Database	Target organisms	Source repository	Method	Sequences	Unique species (%Unique species)	Marine species (%Marine species)	Reference
BOLD	Eukaryotes	BOLD	Keyword search	5,586,934	169,705 (3.04)	18,328 (10.80)	Ratnasingham and Hebert¹⁸
GenBank	Eukaryotes	GenBank	Keyword search	1,933,547	160,061 (8.28)	17,943 (11.21)	NCBI Resource Coordinators²⁰.
Midori	Metazoans	GenBank	Keyword search	927.386	131,988 (14.23)	14,057 (10.65)	Machida, et al.⁴³
db_COI_MBPK	Eukaryotes	EMBL, BOLD	in silico ecoPCR + custom R script	188.975	48,853 (25.85)	6,844 (14.01)	Wangensteen and Turon³⁹
CRUX_CO1	Eukaryotes	EMBL, GenBank	CRUX (in silico ecoPCR + blast)	1,401,802	127,422 (9.10)	15,737 (12.35)	Curd, et al.⁴¹
MARES_BAR	Marine eukaryotes	GenBank, BOLD	Keyword search	1,224,187	61,123 (4.91)	17,884 (29.26)	This data descriptor³⁸
MARES_NOBAR	Marine eukaryotes	GenBank, BOLD	Keyword search	1,491,691	71,499 (4.79)	19,154 (26.79)	This data descriptor³⁸

BOLD and GenBank reference databases were built using Step 1 and 2 of the bioinformatic pipeline (Fig. 1). ‘BOLD’ was generated by retrieving all COI sequences available from the BOLD repository. ‘GenBank’ was generated with the keyword search Eukaryota and COI synonyms. ‘Unique species’ were retained after a quality control procedure that retains only unique, fully identified taxa with binomial species names. ‘% Unique species’ was calculated using the number of unique species as the numerator and the total number of sequences as the denominator. ‘Marine species’ was determined by the number of unique species present in each database that appeared in the World Register of Marine Species (WoRMS) and AlgaeBase²⁷. ‘% Marine species’ was then calculated using the number of marine species as the numerator and the number of unique sequences as the denominator.

Search