Updated TB-Profiler: enhanced genotypic antimicrobial resistance prediction and relatedness analysis powered by a database of over 171,000 Mycobacterium tuberculosis genomes

Abstract

TB-Profiler is a user-friendly bioinformatics tool developed to genotypically profile Mycobacterium tuberculosis from next-generation sequencing data. It provides predictions of drug resistance and assigns sub-lineages to support clinical management of tuberculosis (TB) and public health surveillance. The platform has become widely adopted for genotypic antimicrobial susceptibility prediction, initially covering 16 anti-TB drugs. However, increasing clinical and epidemiological demands have highlighted the need for expanded functionality, including updated resistance mutation libraries, integration of large-scale genomic datasets, and tools to infer genomic relatedness for identifying transmission events and outbreaks. TB-Profiler (v6.6.5) has been extended to address these needs. Resistance mutation libraries have been expanded to include 17 anti-TB drugs, including delamanid and pretomanid, incorporating WHO-endorsed interpretation rules and curated loss-of-function mutations. Supported drugs include bedaquiline and clofazimine, cycloserine/terizidone, and para -aminosalicylic acid. A curated and continuously expanding global M. tuberculosis genomic database comprising more than 170,000 isolates from 136 countries and representing all major lineages has been integrated into the platform. This resource enables allele frequency comparisons within a global population context. The database is linked to new analytical functionality for calculating and visualising genomic relatedness between isolates. We demonstrate its utility by identifying potential transmission events among isolates from Uganda. The usefulness of TB-Profiler for both clinical decision-making and surveillance is further enhanced through multilingual reporting capabilities, facilitating broader accessibility and implementation. These enhancements to TB-Profiler reflect the evolving needs of genomic research and public health practice. Future developments will leverage the expanding sequencing database to implement AI approaches for refining lineage assignment, drug resistance prediction, and transmission classification, including the identification of previously uncharacterised resistance-associated mutations. Stand-alone and web-based versions of TB-Profiler, along with associated databases, are available at https://tbdr.lshtm.ac.uk .