highly significantP < 2.0e−16
In parallel, a linear model assessing the direct association between mean relative abundance and response demonstrated a highly significant positive effect (estimate = 0.30, P < 2.0e−16), with dataset terms again showing no significant influence (all P > 0.05), underscoring the robustness of abundance-based signals across cohorts.