Mycobacterium tuberculosis remains one of the world's most persistent pathogens, responsible for over a million deaths each year despite effective drugs and decades of research. Disease progression varies widely among individuals, reflecting interactions between pathogen, physiology, host immunity, and pharmacological response. We propose a digital twin framework for tuberculosis that integrates clinical, immunological, and pharmacokinetic data into a continuously adaptive computational model. The twin would simulate host-pathogen dynamics from granuloma formation to systemic immune regulation, linking these processes with individualised drug exposure and treatment response. By forecasting outcomes and identifying early indicators of relapse or resistance, such a system could guide precision therapy and accelerate discovery of host-directed interventions. The tuberculosis digital twin thus represents a bridge between infection biology, computation, and clinical translation, presenting an evolving model capable of transforming how this ancient disease is understood and managed. However, translation will require further longitudinal clinical and immunological datasets and systematic validation of model predictions in real-world treatment settings.
Siddiqui et al. (Thu,) studied this question.