My current area of interest lies at the intersection of epidemiology, bioinformatics, computational and evolutionary biology as applied to microbial genome science and more specifically to understanding the biology of drug resistant Mycobacterium tuberculosis.

My previous work has included highly cited work on the accuracy of tuberculin skin tests for the diagnosis of latent tuberculosis, and the development of mathematical metabolic models to understand M. tuberculosis biology and response to drug exposure.

My most recent work has focused on the analysis and interpretation of signatures of natural selection in whole genome sequences of M. tuberculosis to uncover novel genes and cellular mechanisms associated with resistance. Through performing this work I developed a novel interface for storage and access of combined genomic and phenotypic data in Mtb that can significantly decrease time necessary for data retrieval and analysis. Additionally we have applied machine learning techniques to predict the M. tuberculosis resistance phenotype using the genetic sequence data of the established resistance genes in a large set of drug resistant M. tuberculosis strains. In other work we have performed a meta-analysis reviewing mutations in TB that are causative of drug resistance as determined by allelic exchange experiments and summarizing the evidence supporting this. We are also designing a schema for clinical strain selection and statistical power assessment for the prospective design of studies examining genotypic correlated of binary phenotypes in microbial populations, investigating  genomic sequence applications to decipher disease outbreaks of drug resistant M. tuberculosis, and predicting the clinical implications of using novel diagnostics for detection and guidance of TB patient therapy.