This study proposes a novel machine learning (ML) and deep learning (DL) integrated multi-omics approach for leukemia prediction. The research analyzes multi-omics data using various ML and DL algorithms, including Random Forest (RF), Naive Bayes (NB), Decision Tree (DT), Logistic Regression (LR), Gradient Boosting (GB), Recurrent Neural Networks (RNN), and Feedforward Neural Networks (FNN). GB achieved 97% accuracy in ML, while RNN outperformed with 98% accuracy in DL. The study evaluates the effectiveness of these techniques using 17 features such as patient age, sex, mutation type, treatment methods, and chromosomes. The research highlights the importance of DL in leukemia prediction and demonstrates the potential of high-throughput technology in healthcare. The study compares ML and DL techniques and selects the best method for optimal results. The proposed approach integrates multi-omics data to predict leukemia with high accuracy. The study uses a 3-genomics dataset and selects features based on correlation coefficients. The research evaluates the performance of ML and DL classifiers and proposes the most accurate ones for detecting leukemia. The study also discusses the limitations of traditional ML algorithms and the advantages of DL in complex biological research. The results show that RNN achieved the highest accuracy of 98% for binary cross-entropy loss, while GB demonstrated the highest Precision, Recall, and f1 scores of 96%, 98%, and 97%, respectively. The study concludes that the proposed approach has significant potential for improving leukemia prediction and highlights the importance of integrating multi-omics data in healthcare. Future work will explore other omics technologies for predicting different diseases and integrate the findings into practical, real-time scenarios.This study proposes a novel machine learning (ML) and deep learning (DL) integrated multi-omics approach for leukemia prediction. The research analyzes multi-omics data using various ML and DL algorithms, including Random Forest (RF), Naive Bayes (NB), Decision Tree (DT), Logistic Regression (LR), Gradient Boosting (GB), Recurrent Neural Networks (RNN), and Feedforward Neural Networks (FNN). GB achieved 97% accuracy in ML, while RNN outperformed with 98% accuracy in DL. The study evaluates the effectiveness of these techniques using 17 features such as patient age, sex, mutation type, treatment methods, and chromosomes. The research highlights the importance of DL in leukemia prediction and demonstrates the potential of high-throughput technology in healthcare. The study compares ML and DL techniques and selects the best method for optimal results. The proposed approach integrates multi-omics data to predict leukemia with high accuracy. The study uses a 3-genomics dataset and selects features based on correlation coefficients. The research evaluates the performance of ML and DL classifiers and proposes the most accurate ones for detecting leukemia. The study also discusses the limitations of traditional ML algorithms and the advantages of DL in complex biological research. The results show that RNN achieved the highest accuracy of 98% for binary cross-entropy loss, while GB demonstrated the highest Precision, Recall, and f1 scores of 96%, 98%, and 97%, respectively. The study concludes that the proposed approach has significant potential for improving leukemia prediction and highlights the importance of integrating multi-omics data in healthcare. Future work will explore other omics technologies for predicting different diseases and integrate the findings into practical, real-time scenarios.