Dylan Adlard

This study uses a graph convolutional network to predict pyrazinamide resistance in Mycobacterium tuberculosis by integrating protein structure and residue-level features, achieving high accuracy.

This paper develops an automated method to build a comprehensive catalogue of resistance-associated variants for M. tuberculosis, achieving comparable performance to existing catalogues without expert rules.

The paper develops an automated method to predict bedaquiline resistance in M. tuberculosis using a reproducible algorithmic approach, achieving high sensitivity and specificity in variant classification.

This study trains machine learning models to predict rifampicin resistance in M. tuberculosis based on structural and chemical features of rpoB mutations, achieving moderate accuracy.

This paper predicts pyrazinamide resistance in Mycobacterium tuberculosis using machine learning models trained on pncA mutations and structural features, achieving 80.2% sensitivity and 76.9% specificity.