Doxorubicin is an effective anti-cancer drug, but its use is limited by dose-dependent cardiotoxicity, leading to long-term heart damage. This thesis explores the application of machine learning and deep learning models to analyse single-cell RNA sequencing data from cardiomyocytes exposed to doxorubicin. Models found that Drug_D2 samples were easily classified, achieving an AUC of ~0.9, while distinguishing between Control_D2/14 and Drug_D14 samples proved more challenging, with AUCs ranging from ~0.7 to 0.8. Classic ML models, such as Logistic Regression, Random Forest, and XGBoost, showed higher consistency in identifying important features linked to known cardiotoxic pathways compared to artificial neural networks. However, ANNs may uncover previously unexplored features significant for understanding doxorubicin-induced cardiotoxicity. Future work should explore DL models’ types, integrate multi-omics data, and explore unprocessed data to enhance feature extraction, aiming to improve prediction models for patient-specific cardiotoxicity risks and guide new therapeutic strategies.