Abstract: Surveillance of circulating drug resistant bacteria is essential for healthcare providers to deliver effective empiric antibiotic therapy. However, molecular epidemiology does not occur on a timescale that is optimal for guiding patient treatment. Here the Inventors present a method called neighbor typing for inferring characteristics of an unknown bacterial sample by identifying the its closest relative in a database of known genomes. The Inventors demonstrate an implementation of this principle using sequence k-mer content, to identify both the closest relative and a phenotype of interest, in this case drug resistance. The Inventors show for the examples of S. pneumoniae and N. gonorrhoeae that this technique can be applied to data from an Oxford Nanopore device in real time and is capable of identifying the presence of a known resistant strain in 5 minutes of sequencing and 4 hours from sample collection, even from a clinical metagenomic sample.