Calibrating for absolute microbiome abundances without spike-ins

Metagenomics is a widely used approach in microbiome research. However, a major limitation of metagenomic datasets is their compositional nature, which prevents direct quantification of absolute abundances and complicates cross-sample comparisons. Existing strategies for absolute quantification typically require additional experiments or spike-in controls. Here, we introduce MGCalibrator, a new tool that enables spike-in free, absolute abundance estimation based on routine DNA concentration measurements (Qubit). We validated the accuracy of absolute abundances obtained with MGCalibrator against qPCR for 5 targets. Our results show a strong correlation with qPCR data, indicating that MGCalibrator enables qPCR-like trend analyses. For Bacteroides dorei, the estimated abundances were highly similar between the two methods. For crAssphage, MGCalibrator produced estimates approximately 7-fold higher than qPCR, while for 16S rRNA and vanA, it yielded values about 4-fold and 17-fold lower, respectively. Benchmarking with synthetic microbiome data demonstrated that our method accurately determines copy numbers in sequencing datasets, and application to mock community samples confirmed that it does not overestimate abundances. Finally, we applied MGCalibrator to track temporal trends in antibiotic resistance genes (ARGs) in wastewater treatment plants in two Dutch provincial capitals. We observed an overall increase in ARGs—such as sul2 in Utrecht and qnrS5 in Houtrust—likely driven by rising bacterial loads. Our findings demonstrate that MGCalibrator provides robust calibration of metagenomic data, paving the way for metagenomics to play a central role in future surveillance by enabling trend analysis across thousands of genetic targets, similar to the capabilities of qPCR for individual genes.