Supplementary file 1_Longitudinal dynamics of the gut microbiome during methamphetamine addiction and abstinence.docx

Time-series analyses revealed characteristic alterations in Verrucomicrobia and Actinobacteria during prolonged addiction and early abstinence. At the genus level, Butyricimonas, Candidatus Arthromitus, Enterococcus, and Turicibacter exhibited reproducible phase- and temporal-dependent dynamic patterns. Based on these microbial signatures, random forest models were constructed for phase classification and temporal prediction.

The classification model achieved higher model performance during the addiction phase, whereas performance declined during abstinence, reflected by reduced variance explained and increased heterogeneity in gut microbiota composition. Temporal prediction models further demonstrated differential predictive performance between phases.</p>Discussion<p>These findings identify reproducible temporal changes in gut microbiota across addiction and abstinence and highlight candidate microbial signatures for understanding addiction progression and distinguishing different phases.</p>