Shotgun metagenomic assessment of raw and treated wastewater from a public oncology hospital in Paraguay : treatment-associated shifts in DNA-based bacterial profiles and resistome signatures

Abstract

Hospital wastewater is an important matrix for monitoring clinically relevant bacteria and antimicrobial resistance genes (ARGs). We used shotgun metagenomic sequencing for an exploratory comparison of untreated wastewater and treated effluent from a public oncology hospital in Paraguay during four sampling campaigns (2024–2025). Four campaign-matched pairs generated eight metagenomic datasets. Taxonomic profiles were inferred using Kraken2/Bracken, alpha and beta diversity were assessed, selected clinically relevant bacterial genera were examined, and ARG-associated sequences were identified in assembled contigs using CARD. Rarefaction analysis indicated adequate genus-level sequencing coverage. Treated effluents showed a higher relative representation of Pseudomonadota and lower representation of Bacillota and Bacteroidota. Paired comparisons of Shannon and Simpson diversity did not detect statistically significant differences between effluent types, although both showed large effect sizes (Cohen’s dz). Similarly, the campaign-blocked PERMANOVA showed no statistically significant difference in genus-level bacterial composition based on Bray–Curtis dissimilarity, despite a marked effect size. Clinically relevant bacterial genera were detected in both sample types, with Pseudomonas predominating in treated wastewater. Treated effluents also contained markedly fewer contig-based ARG detections and lower sequencing-depth-normalized ARG richness across all matched comparisons. The results suggest treatment-associated differences in bacterial profiles and fewer detectable ARGs, providing baseline data for hospital wastewater surveillance in Paraguay.