Global environmental genomics reveals vast uncultivated protist diversity and taxonomic blind spots

Abstract

Protists, defined as eukaryotes distinct from plants, animals and fungi, comprise much of eukaryotic phylogenetic diversity and perform key ecological functions, yet most lineages remain uncultivated and poorly represented in reference databases. Amplicon-based 18S rRNA gene surveys have transformed protist ecology, but primer bias, PCR bias and marker length can limit recovery of divergent lineages. Here, we conduct a cultivation-independent, rDNA operon-centered survey of over 30,000 assembled metagenomes, protist single-cell genomes and 21 long-amplicon datasets spanning marine, freshwater and soil-associated environments. We recover 157,956 long 18S rRNA gene sequences (≥800 bp), which together with reference sequences cluster into 103,338 OTUs at 97% identity and 24,438 OTUs at 85% identity. After excluding non-protist groups and singletons, 81% of the remaining 9543 protist 85%-OTUs consist exclusively of environmental sequences with no described representative. A concatenated 18S-28S rRNA framework reveals strong underrepresentation of Amoebozoa, Discoba, and Rhizaria in current references and identifies 13 candidate deeply divergent lineages with little or no database representation, some of which we link to partial genomes recovered by single-cell genomics. We further characterize their ecosystem distributions and predict associations with bacteria and other eukaryotes. This study provides a global, cultivation-independent resource for targeted sampling and downstream genomic, ecological and experimental characterization. In this study, the authors analyze over 30,000 assembled metagenomes across ecosystems to uncover hidden protist diversity, including major novel uncultivated lineages, and reveal their habitats and connections to other microbes.