Related Experiment Video
Updated: Jun 19, 2025
06:50
Author Spotlight: A Computational Approach to Decipher Amino Acid Preferences in Multispecific Protein-Protein Interactions
Published on: January 26, 2024
1.8K
Prediction of leukemia peptides using convolutional neural network and protein compositions
Seher Ansar Khawaja1, Muhammad Shoaib Farooq1, Kashif Ishaq1
1School of System and Technology, University of Management and Technology, Lahore, 54000, Pakistan.
BMC Cancer
|July 26, 2024
Summary
This study introduces a deep convolutional neural network (CNN) for leukemia detection using protein expression. The CNN model achieved high accuracy, outperforming other methods in identifying leukemia patterns.
Area of Science:
- Biomedical Engineering
- Computational Biology
- Oncology
Background:
- Leukemia is a blood cell cancer originating in bone marrow, with acute and chronic forms further classified into lymphocytic and myeloid types.
- Accurate and early detection of leukemia is crucial for effective treatment and patient outcomes.
Purpose of the Study:
- To evaluate a deep convolutional neural network (CNN) classifier for improved leukemia identification precision using protein expression data.
- To compare the performance of CNN against other machine learning models, including Multilayer Perceptron (MLP), Gated Recurrent Unit (GRU), and Recurrent Neural Network (RNN).
Main Methods:
- Utilized protein expression data from leukemia patients for classification tasks.
- Employed various feature extraction techniques, including Amino Acid Composition (AAC), Group of AAC (GAAC), Tripeptide Composition (TPC), Group of TPC (GTPC), and Dipeptide Composition (DPC).
- Implemented and validated models using independence and cross-validation testing, analyzing Receiver Operating Characteristic (ROC) curves.
Main Results:
- The CNN model demonstrated superior performance compared to MLP, GRU, and RNN in both independence and cross-validation tests.
- In independence testing with AAC and GAAC features, CNN achieved 100% accuracy.
- Cross-validation testing using AAC and GAAC features yielded a highest accuracy of 98.33% for the CNN model. GRU achieved 0.965% accuracy on ROC curves.
Conclusions:
- The developed CNN model shows significant potential for accurate leukemia detection from protein expression data.
- The findings highlight the effectiveness of deep learning approaches, particularly CNNs, in analyzing complex biological data for cancer diagnostics.
