Effects of host DNA and k-mer size on metagenomic classification of historical samples

Microbiome composition and function are strongly influenced by environmental factors, with major shifts driven by intensified anthropogenic pressures over the past centuries. This timeframe extends beyond the scope of experimental and longitudinal studies traditionally used to investigate microbiome dynamics. Historical samples could provide important insights into the mechanistic consequences of anthropogenic pressures and the resulting shifts in microbial diversity and composition. Despite their vast potential, historical samples held in museums and herbaria worldwide remain underutilized for exploring host-microbiome interactions across broad temporal and spatial scales, owing to their incompatibilities with standard analytical pipelines and a limited understanding of optimal analytical parameters. Host DNA removal has conventionally been considered essential for taxonomic assignment of metagenomic reads, and may be particularly important when processing degraded DNA, but it is impractical for specimens whose host species lacks a reference genome. Here, using empirical and simulation datasets, we show that host DNA content has a negligible impact on microbial data analysis. Because DNA molecules from historical samples are highly fragmented and variable in length, we further analysed the effect of k-mer size on the classification of metagenomic reads from historical samples. To improve recall rate, we propose a simple two-step approach in which reads are classified against two databases constructed with long and short k-mer sizes. Using simulated and published datasets, we demonstrated that this approach outperforms single-step workflows in recovering microbial signals from reads across a wide range of lengths. Together, our findings provide a foundation for incorporating natural history collections into host-associated microbiome research, offering insights into the long-term effects of anthropogenic change on microbial communities.