The optimization of gene sequencing techniques over the last few decades enables the efficient identification of genetic mutations associated with disease. However, this efficiency now presents scientists with a data overload problem, as identified variation cannot be as efficiently confirmed causative through traditional molecular techniques. To address this, I have developed a structural bioinformatics pipeline which analyses the effects of confirmed causative missense mutations and uses that insight to predict the phenotype of novel mutations encountered in the clinic.
By characterizing mutations within their 3-dimensional protein structure through estimations of stability and affinity changes, I was able to elucidate distinct mechanisms of tuberculosis drug resistance across three first- and second-line drug targets/activators alr, katG and rpoB. In doing so, I found that resistance was mediated mostly through changes in non-ligand interactions across the three genes. I later adopted a similar approach to decipher the molecular drivers in amyotrophic lateral sclerosis, and found that destabilization, which is a precursor to protein aggregation, is not the main driver of disease across highly mutated genes SOD1, FUS and TDP-43.
Using similar approaches and insights within a machine learning pipeline, I developed a predictive tool for detecting antimicrobial resistance to rifampicin, which detects resistance across the whole gene rpoB, and consistently outperformed the FDA-endorsed gold standard GeneXpert-MTB-RIF, and is crucial for treating tuberculosis and leprosy infections. In applying the same method to distinguish autism and cancer phenotypes in the human gene PTEN, I developed the first-of-its-kind multiclass predictor, which has important clinical utility within PTEN hamartoma tumour syndrome, as it can direct patient treatment protocols. My overall approach in analysing mutations has proven direct clinical applications, while also providing insights into the molecular mechanisms of disease in a high-throughput manner, which are invaluable for development of novel therapeutic strategies, particularly in the era of precision medicine.