Abstract
Long-read sequencing has revolutionised metagenomics through improved metagenome assembly, taxonomic classification and functional characterisation. Automation can enhance the throughput, reproducibility, and accuracy of library preparation. However, the validation of automated library preparation protocols remains undetermined for metagenomic workflows, which are particularly sensitive to methodological perturbation. Here, we compare long-read metagenomic sequencing of environmental samples through parallel manual and automated protocols. Despite minor variation in read lengths and classification rate, minimal differences in microbial community structure were identified between manual and automated libraries. These findings demonstrate the utility of automation for high-throughput long-read metagenomics, with broad applicability to automated long-read sequencing.