Hologenomic data generation and analysis in wild vertebrates

Abstract

Hologenomics, the joint analysis of host genomes and microbial metagenomes, has the potential to address fundamental biological questions from a systemic host‐microbiota perspective. However, multiple fieldwork, laboratory and bioinformatic steps challenge quality, representativeness and comparability of hologenomic data. Leveraging the first 2025 samples sourced from 151 wild vertebrate species analysed in the Earth Hologenome Initiative, we scrutinise hologenomic data generation steps, including laboratory and bioinformatic procedures. Comparisons across taxa and sample types provide novel insights into the relationships between laboratory quality metrics and derived data, the variation of host, prokaryotic and non‐prokaryotic fractions of shotgun data, and the relationship between data quality and quantity with genome and metagenome reconstruction. Our results show that faecal samples are significantly better than anal and cloacal swabs to study intestinal microbiomes using genome‐resolved metagenomics. We also report that birds and bats both have substantially lower microbial DNA fractions and a higher degree of sample‐to‐sample variability compared to amphibians, reptiles and non‐flying mammals. Based on these data, we provide suggestions to the field for robustly and efficiently generating hologenomic data from wild vertebrates.