Αlpha¹ Review in Progress

The detection of low-biomass microbial DNA in human tissues is often confounded by contamination, as demonstrated in the debates over the existence of microbiomes in the placenta, brain, blood, and tumors. Here we show that genomic DNA fragment length serves as an informative discriminator: while genuine microbiome genomes have long genomic DNA fragments, contaminant DNA is typically short and fragmented. Using germ-free mouse tissues with bacterial spike-ins and human cell lines, we developed Median Length-Adjusted (Median(L)adj), a metric that normalizes microbial read length to host read length, facilitating the differentiation between genuine microbiome and contamination. Applying the metric across multiple human tumor and normal tissues, we found genuine microbiome signals are largely limited to biopsy sites with natural microbial exposure (e.g., gastrointestinal tract, cervix, vagina and skin), while sterile tissues, including kidney, brain, lung, blood, and placenta, showed no evidence of resident microbiome. These findings support DNA fragment length as an informative metric for quality controlling low-biomass microbiome profiling, offering a framework to clarify the ongoing debates and strengthen future studies of resident microbiome in tissues.