Predicting antibiotic resistance in complex protein targets using alchemical free energy methods

Multi-drug resistant Mycobacterium tuberculosis requires a complex antibiotic treatment program and poses a major threat to tuberculosis (TB) treatment outcomes. Resistance is mostly conferred by chromosomal single nucleotide polymorphisms, many of which are well characterized and catalogued. However, not all mutations have been mapped and novel mutations can emerge. Methods able to quickly predict the effects of such mutations are needed to complement the existing catalogues, thereby permitting the prescription of effective treatment for patients and preventing the further spread of resistant strains. Relative binding free energy (RBFE) calculations can rapidly predict the effects of mutations, but this approach has not been tested on large, complex proteins. We use RBFE calculations to predict the effects of seven M. tuberculosis RNA polymerase mutations on rifampicin susceptibility and five M. tuberculosis DNA gyrase mutations on moxifloxacin susceptibility. These mutations encompass a range of amino acid substitutions with known effects and include large steric perturbations and charged moieties. We find that moderate numbers (n=3-15) of short RBFE calculations can predict resistance in cases where the mutation results in a large change in the binding free energy, but that the method lacks discrimination in cases with either a small change in energy or that involve charged amino acids, due to the associated large magnitude of error. We investigate how this error may be decreased by analyzing the sources of error and the distributions of repeated measurements from the different components of the RBFE calculations.