Resistome Changes During Dental Development

Differential abundance analysis (DAA) was used to assess the impact of time point or stage of dental development and caries at T3 on the resistome. Prior to DAA, the data were filtered by removing low abundance ARGs (cutoff: 0.01% counts) and ARGs present in only one sample. To assess the impact of time point, we used a linear mixed effects model (MaAslin2), where time point was treated as a fixed effect and individual as a random effect, to account for repeated sampling per person. Analysis was performed on ARGs present in at least 20% of the samples that had been normalised by total sum scaling (TSS) and log transformed. To ensure robust biological interpretation^{75 (link)}, we used two DAA methods (MaAslin2 v1.0.0 and ANCOM-BC v2.1.2), to investigate the impact of caries (mild, moderate, severe) on ARGs at T3. We then assessed if treating caries (restorations) impacted ARGs where caries status (caries free or affected) and restorative status (restored [yes] or unrestored [no]) were treated as fixed effects. Default parameters were used in both packages; however, alpha was adjusted to 0.25 in ANCOM-BC to be consistent with MaAsLin2.