Understanding the complex relationships between medicinal drugs is of paramount importance and is a challenge with the rapid development of new drugs. In this paper Graph Neural Networks (GNNs) are used to analyze drugs in a new way. Realizing that drug molecules naturally look like graphs, with atoms as nodes and bonds as lines, this work uses the power of GNNs to get both local and global information from the molecular structures. GNN analyzes complex patterns and relationships, which gives more information about how drugs interact with each other, what side effects they might have, and how they work. Comparing GNN-based approach to traditional ways of analyzing drugs, it is observed that GNN networks analyses is better at predicting drug properties, drug-drug interactions, and chances for repurposing drugs. This study demonstrates how GNNs could change the way drug analysis is done and also paves the way for more thorough, efficient, and accurate ways to find new drugs in the future.