Classification of cancer types based on microRNA expression using a hybrid radial basis function and particle swarm...

Area of Science:

• Biotechnology
• Bioinformatics
• Genomics

Background:

• MicroRNAs (miRNAs) are crucial regulators of gene expression implicated in various cancers.
• Accurate cancer-type classification from miRNA expression data is vital for effective diagnosis and treatment.
• Current methods face challenges in accuracy and computational efficiency.

Purpose of the Study:

• To develop an improved computational method for accurate cancer-type classification using miRNA expression data.
• To enhance classification accuracy and reduce computational load through optimized feature selection.

Main Methods:

• A hybrid approach combining radial basis function (RBF) neural networks with particle swarm optimization (PSO) was employed.
• PSO was utilized for efficient feature selection to identify the most relevant miRNA biomarkers.
• Data preprocessing and normalization were performed on two distinct miRNA expression datasets.

Main Results:

• The proposed method achieved high classification accuracies of 95% and 91% on the two datasets.
• An average accuracy of 93% was obtained, demonstrating superior performance compared to existing methods.
• PSO effectively reduced the feature set, contributing to computational efficiency.

Conclusions:

• The hybrid RBF-PSO method offers a robust and accurate approach for cancer-type classification based on miRNA expression.
• This technique holds significant potential for advancing cancer diagnostics and personalized medicine.
• Minimal feature selection via PSO combined with RBF classification improves reliability and efficiency.