A machine learning and deep learning-based integrated multi-omics technique for leukemia prediction

Abstract

In recent years, scientific data on cancer has expanded, providing potential for a better understanding of malignancies and improved tailored care. Advances in Artificial Intelligence (AI) processing power and algorithmic development position Machine Learning (ML) and Deep Learning (DL) as crucial players in predicting Leukemia, a blood cancer, using integrated multi-omics technology. However, realizing these goals demands novel approaches to harness this data deluge. This study introduces a novel Leukemia diagnosis approach, analyzing multi-omics data for accuracy using ML and DL algorithms. ML techniques, including Random Forest (RF), Naive Bayes (NB), Decision Tree (DT), Logistic Regression (LR), Gradient Boosting (GB), and DL methods such as Recurrent Neural Networks (RNN) and Feedforward Neural Networks (FNN) are compared. GB achieved 97 % accuracy in ML, while RNN outperformed by achieving 98 % accuracy in DL. This approach filters unclassified data effectively, demonstrating the significance of DL for leukemia prediction. The testing validation was based on 17 different features such as patient age, sex, mutation type, treatment methods, chromosomes, and others. Our study compares ML and DL techniques and chooses the best technique that gives optimum results. The study emphasizes the implications of high-throughput technology in healthcare, offering improved patient care.

Keywords: Deep learning; Genomics; Leukemia; Machine learning; Multi-omics.

Fig. 1

Overview of omics technologies.

Fig. 2

Example of artificial neural networks.

Fig. 3

A. Correlation coefficient 0 B. Correlation coefficient −1.

Fig. 4

Correlation Coefficient with a positive relation.

Fig. 5

Proposed methodology.

Fig. 6

Flow chart of research study.

Fig. 7

A. Correlation coefficient (part 1). B. Correlation coefficient (part 2).

Fig. 7

A. Correlation coefficient (part 1). B. Correlation coefficient (part 2).

Fig. 8

A. Consolidated features (part 1). B. Consolidated features (part 2). C. consolidated features (part 3).

Fig. 8

A. Consolidated features (part 1). B. Consolidated features (part 2). C. consolidated features (part 3).

Fig. 8

A. Consolidated features (part 1). B. Consolidated features (part 2). C. consolidated features (part 3).

Fig. 9

Combined ROC curves for classification models.

Fig. 10

A. Confusion matrix for GB; B. Confusion matrix for NB.

Fig. 11

A. Confusion matrix for RF; B. Confusion matrix for DT.

Fig. 12

Confusion matrix for LR.

Fig. 13

A. Mse loss for FNN; B. Bce loss for FNN.

Fig. 14

A. Bce loss for RNN; B. Mse loss for RNN.

Fig. 15

A. Bce loss for FNN; B. Mse loss for FNN.

Fig. 16

A. Mse loss for RNN; B. Bce loss for RNN.

Fig. 17

A. RNN with MSE using Softmax Function; B. RNN with BCE using Softmax Function.

Fig. 18

A. FNN with BCE using Softmax Function; B. FNN with MSE using Softmax Function.

Fig. 19

A. FNN with MSE using Relu Function; B. FNN BCE Loss using Relu Function.

Fig. 20

A. RNN with MSE using Relu Activation Function; B. RNN with BCE using Relu Activation Function.

Fig. 21

Overall performance of DNNs.

Fig. 22

Comparison of ML and DL techniques.