Neglected tropical diseases (NTDs) are a significant health burden globally, contributing to 1–2% of all mortality. The quest for new pharmacological interventions for NTDs is hampered by the substantial time and financial resources required for drug development, alongside a marked disinterest from laboratories due to the geographic and economic profiles of affected regions. This study presents a novel protocol for leveraging graph database architectures to elucidate potential connections between NTDs and existing pharmaceuticals that may be repurposed for treatment. Focusing on tuberculosis as a case study, we constructed a graph database integrating comprehensive disease and drug data. By removing established links between tuberculosis and known treatments, we employed a Graph Convolutional Neural Network (Graph-CNN) algorithm to unearth previously unrecognized relationships. This computational approach not only facilitates the discovery of new drug-disease associations but also streamlines the repurposing process for existing medications. Our findings demonstrate the utility of advanced data engineering and machine learning techniques in identifying promising therapeutic candidates for NTDs, potentially accelerating the availability of effective treatments for these underserved diseases. The study’s contribution is underscored by a comparative analysis of state-of-the-art algorithms within our graph database framework, highlighting the efficacy of our protocol in the broader context of drug repurposing research.