Abstract
This study leverages the Bacterial and Viral Bioinformatics Resource Center (BV-BRC) to analyze over 27,000 Mycobacterium tuberculosis (MTB) genomic strains, providing a comprehensive and large-scale overview of antibiotic resistance (AMR) prevalence and resistance patterns. We used MTB++, which is the newest and most comprehensive AI-based MTB drug resistance profiler tool, to predict the resistance profile of each of the 27,000 MTB isolates and then used feature analysis to identify key genes that were associated with the resistance. There are three main contributions to this study. Firstly, it provides a detailed picture of the prevalence of specific AMR genes in the BV-BRC dataset as well as their biological implications, providing critical insight into MTB's resistance mechanisms that can help identify genes of high priority for further investigation. The second aspect of this study is to compare the prevalence of antibiotic resistance across previous studies that have addressed both the temporal and geographical evolution of MTB drug resistance. Lastly, this study emphasizes the need for targeted diagnostics and personalized treatment plans. In addition to these contributions, the study acknowledges the limitations of computational prediction and recommends future experimental validation.