The scope of this project is to investigate dynamic and persistent proteomic changes in tuberculosis (TB) patients during the course of treatment by analyzing paired bronchial alveolar lavage fluid (BALf) and serum samples collected at diagnosis and after 25 weeks of therapy. By comparing patient samples to those from healthy community controls, the study aims to identify protein signatures that reflect both local (pulmonary) and systemic host responses, providing insight into treatment-associated changes and the potential risk of disease recurrence. We developed and applied a sample-specific processing workflow that includes bacterial inactivation steps, ensuring compatibility with downstream applications such as metabolomics analysis. Proteomic data were acquired using a data-independent acquisition (DIA) approach with a library-free analysis strategy, followed by a robust two-step imputation method to address missing values. The resulting high-confidence proteome datasets offer a valuable resource for future comparative analyses and biomarker discovery.