New PreMiEr Publication: New Statistical Model Improves Analysis of Microbiome Data Across Body Sites
PreMiEr researchers have published a new statistical model in BMC Bioinformatics that improves microbiome analysis across paired body sites by capturing cross-site biological dependencies that standard methods miss.
A new statistical tool developed by PreMiEr researchers improves the analysis of microbiome sequencing data collected from paired body sites, such as the bladder and vagina, by accounting for the biological connections between them. Published in BMC Bioinformatics, the study introduces a Bayesian negative binomial latent factor model that captures cross-site dependencies often missed by standard analytical approaches. In simulations and a real-world case study of the female urogenital microbiome, the model outperformed existing methods in detecting biological effects, predicting outcomes, and revealing differences among individuals in how their microbial communities interact across body sites. The framework offers a more accurate picture of how microbiomes function in interconnected environments throughout the body.
Read more here: https://doi.org/10.1186/s12859-025-06362-3