Pharmacogenomics using deep learning is an emerging research area that combines genomics, pharmacology, and artificial intelligence to understand how genetic variations influence drug response and efficacy. Early studies applied traditional machine learning models to predict gene-drug interactions and adverse drug reactions, while recent research leverages deep learning architectures such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), graph neural networks (GNNs), and attention-based models to capture complex patterns in genomic sequences, molecular structures, and multi-omics data. Applications include drug response prediction, personalized medicine, biomarker discovery, adverse drug reaction prediction, and drug repurposing. Recent studies also explore integrating heterogeneous datasets, such as genomic, transcriptomic, proteomic, and electronic health record data, to improve model generalization and predictive accuracy. Deep learning-based pharmacogenomics provides scalable and data-driven approaches for precision medicine, enabling more accurate, individualized therapeutic strategies and contributing to safer and more effective drug development pipelines.