RETRACTED ARTICLE: Ensemble recurrent neural network with whale optimization algorithm-based DNA sequence...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Biomedical and clinical data collection has surged in the data-driven era.
• Deoxyribonucleic acid (DNA) gene expression datasets are crucial for identifying pathogens via biomarkers.
• Metaheuristic-related feature selection (FS) is vital for managing large gene datasets.

Purpose of the Study:

• To apply the whale optimization algorithm (WOA) for feature selection in high-dimensional (HD) microarray data.
• To develop an ensemble recurrent neural network (ERNN) for classifying selected gene expression data.
• To evaluate the ERNN's performance against existing advanced methodologies for pathogen detection.

Main Methods:

• Utilized the whale optimization algorithm (WOA) for efficient feature selection from HD microarray datasets.
• Developed an ensemble recurrent neural network (ERNN) integrating long short-term memory (LSTM), bidirectional LSTM, and gated recurrent units (GRUs).
• Classified the selected gene features using the proposed ERNN model.

Main Results:

• The WOA effectively filtered pertinent genes from large feature sets, reducing computational load.
• The ERNN model achieved high performance in classifying gene expression data.
• The proposed ERNN methodology attained 99.59% precision and 99.59% accuracy.

Conclusions:

• The whale optimization algorithm is effective for feature selection in gene expression data analysis.
• The ensemble recurrent neural network demonstrates superior performance for pathogen detection using genomic biomarkers.
• This integrated approach offers a promising strategy for accurate and efficient analysis of biomedical data.