P50

Non-tuberculous mycobacteria (NTMs) are ubiquitous environmental microbes and are opportunistic pathogens which often cause disease in patients with pre-existing conditions. Ascertainment of species and drug susceptibility profile is vital for clinical decision making. Building on our globally used TB-Profiler software tool for Mycobacterium tuberculosis, we developed NTM-Profiler, an end-to-end solution that analyses whole genome sequencing data to perform speciation and predict drug resistance for NTMs. Using a collection of >3k genomes, species-specific short nucleotide sequences (k-mers) were characterised across 221 species, allowing for new samples to be speciated rapidly, direct from raw sequence data. We characterised genetic determinants of resistance (macrolides, amikacin, and rifampicin) for two clinically relevant NTMs (M. abscessus and M. kansasii). The performance of NTM-Profiler was evaluated using a published M. abscessus dataset (N=1,729), and predicted sub-species correctly for all isolates, with high concordance with known resistance profiles across three different subspecies tested. In conclusion, NTM-Profiler is a rapid and robust tool that can predict species and drug resistance. As more public health authorities move towards the use of sequencing to assist diagnosis, our tool is well placed to assist treatment decision making and aid the implementation of personalised medicine.