Functional annotation of eukaryotic genes from sedimentary ancient DNA

Ugur Cabuk ^1,2 Ulrike Herzschuh ^1,2,3 Lars Harms ⁴ Barbara von Hippel ¹ Kathleen Stoof-Leichsenring ^1*

• ¹ Polar Terrestrial Environmental Systems, Department of Earth Sciences, Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research (AWI), Potsdam, Brandenburg, Germany
• ² Institute for Biochemistry and Biology, University of Potsdam, Potsdam, Brandenburg, Germany
• ³ Institute of Enviornmental Science and Geography, Faculty of Mathematics and Natural Sciences, University of Potsdam, Potsdam, Germany
• ⁴ Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research, Computing and Data Centre, Bremerhaven, Germany

Sedimentary ancient DNA (sedaDNA) provides valuable insights into past ecosystems, yet its functional diversity has remained unexplored due to potential limitations in gene annotation for short-read data. Eukaryotes, especially, are typically underrepresented and have low coverage in complex metagenomic datasets from sediments. In this study, we evaluate the potential of eukaryotic gene annotation in sedimentary ancient DNA (sedaDNA) time-series data covering the last 23,000 years. We compared four gene annotation pipelines (GAPs) that apply Prodigal (ProkGAP) and MetaEuk (EukGAP) with and without taxonomic pre-classification. We identify ProkGAP as the pipeline which recovers the largest gene catalog with 6,568,483 functional genes and the highest number of eukaryotic functional genes (5,895 unique KEGG orthologs). Our findings show that ProkGAP, originally invented for prokaryotic gene prediction, yields the largest share of functional genes among all GAPs tested. At the same time, it allows the analysis of prokaryotic and eukaryotic gene functions in parallel and predicts most gene diversity. Interestingly, in our time-series data the gene catalog size and diversity show an increasing trend towards recent times indicating a more complex eukaryotic functional community during the Holocene. However, all gene annotation pipelines are limited by incomplete functional reference databases, which hamper the link between taxonomic-functional relationships when considering lower taxonomic levels. Future research on functional gene prediction from short read sedaDNA data should focus on expanding the eukaryotic databases and increasing sequencing depth to explore eukaryotic and prokaryotic functional composition and diversity in past ecosystems and their relationships to environmental change.