Probiotic ratio prediction method and system based on multi-source data features

Abstract

This invention discloses a method and system for predicting probiotic ratios based on multi-source data features, belonging to the field of microbial technology. The specific steps of the method are as follows: First, select strains to obtain multi-dimensional data such as genomes, and construct a four-dimensional dataset after standardization; then, mine allergy biomarkers to form a high-sensitivity set; next, construct an efficacy ratio model and output strain rankings; then, verify and screen highly active proprietary strains and culture and preserve them; finally, develop compositions, determine the optimal ratio, use encapsulation technology, and complete the full-chain safety and efficacy verification through multi-stage evaluation. This invention promotes the data-driven evaluation of probiotic efficacy by integrating multi-dimensional data resources, mining allergy biomarkers, constructing models to accurately screen highly active strains, and establishing a strain verification system, optimizing culture media and fermentation processes, scientifically compounding proprietary strains and prebiotics, and completing the full-chain verification of compositions, providing complete technical support for the development of anti-allergy probiotic products.

Images