Dereplication-assisted culturomics enables strain-level ecological analysis of the human gut microbiome

Recent advances in culturomics have enabled large-scale recovery of microbial isolates from the human gut, generating extensive culture collections that bridge metagenomic predictions and experimental validation. However, these isolate resources remain largely underutilized, as conventional culturomics prioritizes the discovery of novel species while massive collections of commensal isolates persist as unexplored biological datasets. Dereplication, particularly based on MALDI-TOF MS spectral features, has been largely regarded as a logistical tool for managing redundancy rather than an analytical asset. Here, we reposition dereplication as an analytical framework for interpreting large-scale culturomics datasets and resolving strain-level ecological patterns. We applied the SPeDE pipeline to a comprehensive collection of 2,231 isolates, including Bifidobacterium spp. and Enterococcus faecium, recovered from healthy donor feces. Spectrum-derived operational isolation units (OIUs) revealed host-associated strain-level repertoires and lineage-like clustering within species. Notably, distinct spectral clusters observed in E. faecium corresponded to clade-level patterns identified through shotgun metagenomic analysis. These findings demonstrate that dereplication-assisted culturomics can extend beyond redundancy control to enable high-resolution ecological interpretation of cultured microbiome datasets. By reframing dereplication as a bridge between large-scale isolate generation and strain-level microbiome ecology, this study outlines a conceptual and practical direction for the next phase of human microbiome research in the post-culturomics era.