A BRIGHT platform advances integrated hazard indexing toward actionable control of human pathogenic bacteria in residential buildings

Human pathogenic bacteria (HPB) in residential buildings pose a persistent threat to resident health, yet current approaches rarely identify representative HPB for monitoring or actionable control guidance. Here, we develop a gene-resolved framework that identifies representative HPB by integrating accessibility (prevalence and abundance) and pathogenicity (virulence, disease-related KEGG pathway, antimicrobial resistance, and gene mobility) into a unified relative hazard potential index. To enable practical use, we implement this framework in an open-source platform, the Bacterial Risk Identification and Gene-based Health Toolkit (BRIGHT), which integrates 25 analytical tools and databases to support application-oriented HPB identification and linkage to environmental drivers. Applying BRIGHT to residential metagenomes collected from ten cities in China across summer and winter, we identified 28 potential HPB; Staphylococcus aureus showed the highest hazard index and was therefore selected as the representative target. By associating S. aureus abundance with environmental parameters, machine learning analysis identified outdoor PM_2.5 as the dominant determinant of potential HPB hazard (R² = 0.37, MAE = 0.021). Partial dependence analysis further suggested outdoor PM_2.5 below ∼28 μg/m³, indoor temperature below ∼25.1 °C, and an air change rate above ∼1.0 h⁻¹. Under these conditions, residents showed a lower reported proportion of potential disease symptoms, at 37%. Overall, BRIGHT offers an application-oriented pathway from metagenomic profiling to decision-ready HPB control, including actionable thresholds for residential built environments.