* 일시 : 5/21(목) 16:30 –
* 연사 : 김용규 대표(바이옴에이츠)
* 장소 : 벤처관 711호
* 제목 : From Microbial Interactions to Translational Products: AI-Guided Design of Defined Consortia via the BASyMCo Platform
* 초록 :
The human microbiome represents a highly complex ecological system in which therapeutic
function emerges not from individual strains, but from dynamic and context-dependent microbial
interactions. However, the lack of systematic frameworks to decode and harness these
interactions has limited the development of reproducible and effective microbiome-based
interventions.
BASyMCo (Bank–Lab–Solver) is an AI-driven platform designed to enable the rational design
of defined microbial consortia. The platform integrates a large-scale repository of cultured
human-derived microbes (Bank), high-throughput co-cultivation and multi-omics datasets
capturing interspecies interactions (Lab), and an explainable machine learning engine that
models and predicts interaction-driven functional outcomes (Solver). By quantifying pairwise
and higher-order interaction effects, BASyMCo enables the identification of synergistic “mate”
strains that enhance engraftment, stability, and functional efficacy of target microbes.
This framework has been translated into product development across multiple domains. In
dermatology, a defined microbial consortium targeting Staphylococcus aureus, a key driver of
atopic dermatitis, demonstrates synergistic antimicrobial and immunomodulatory effects. In
parallel, a consumer-facing application, “Hooago,” has been developed as a functional
formulation designed to mitigate capsaicin-induced discomfort through a multi-modal
mechanism combining physical sequestration and microbiome-mediated modulation.
Together, these results illustrate a scalable approach to move beyond empirical microbiome
interventions toward predictable, mechanism-based design of microbial therapeutics and
functional products.