Antimicrobial resistance (AMR) is a major global public health concern mainly affecting low- and middle-income countries (LMICs) due to lack of awareness, inadequate healthcare and sanitation infrastructure, and other environmental factors. In this study, we aimed to link microbial assembly and covariates (body mass index, smoking, and use of antibiotics) to gut microbiome structure and correlate the predictive antimicrobial gene prevalence (piARG) using PICRUSt2. We examined the gastrointestinal and oral microbial profiles of healthy adults in Pakistan through 16S rRNA gene sequencing with a focus on different ethnicities, antibiotic usage, drinking water type, smoking, and other demographic measures. We then utilised a suite of innovative statistical tools, driven by numerical ecology and machine learning, to address the above aims. We observed that drinking tap water was the main contributor to increased potential AMR signatures in the Pakistani cohort compared to other factors considered. Microbial niche breadth analysis highlighted an aberrant gut microbial signature of smokers with increased age. Moreover, covariates such as smoking and age impact the human microbial community structure in this Pakistani cohort.