Fig. 3

Community structure of cell-free microbial DNA in plasma and patient-matching stool and saliva. A Taxonomic profiles (through 16S rRNA gene sequencing) of patient-matching plasma (batches A and B only), stool and saliva samples and respective DENCs at the phylum level. The top 15 most abundant phyla are shown. B Non-metric multidimensional scaling (nMDS) of pairwise Bray-Curtis dissimilarities calculated from microbial community profiles for patient-matching plasma (batches A and B only), stool and saliva samples and respective DENCs at the ASV-level. C Dendrogram based on hierarchical clustering analysis (UPGMA) of pairwise Bray-Curtis similarities for all plasma samples (batches A–E) and their corresponding DENCs at the ASV-level highlighting batch effects due to sequencing run and DNA extraction batch (DEB). For both, B and C, prior to calculating Bray-Curtis dissimilarities, ASV counts were rarified (i.e. randomly subsampled to the minimum sample size) and square-root transformed. D Alpha diversity measurements of all plasma samples and their corresponding DENCs across the five DNA extraction batches A–E based on the number of observed ASVs (richness) and inverse Simpson’s Index (diversity). ‘+’ p between 0.1–0.05, * p < 0.05, ** p < 0.01, ns-not significant