Step 1: Literature review and research gaps

[danymukesha]

I revised the recent literature (2023–2025) individualizing several gaps in genetic algorithm applications for genomic data:

• Multi-objective optimization: Most GA frameworks focus on single-objective optimization (e.g., minimizing fitness based on gene expression differences). Multi-objective optimization (e.g., balancing accuracy and interpretability) is increasingly relevant for genomic data analysis (Li et al., 2023, Bioinformatics).
• Scalability: High-throughput genomic datasets (e.g., single-cell RNA-seq) require GAs that handle large-scale data efficiently (Wang et al., 2024, Nature Computational Science).
• Interpretability: GAs often produce "black-box" solutions. Integrating domain-specific knowledge (e.g., gene networks) can improve interpretability (Zhang et al., 2025, Genomics).
• Parallelization: Parallel computing is underutilized in many GA implementations, limiting performance on large datasets (Smith & Johnson, 2024, Journal of Computational Biology).
• Dynamic adaptation: Adaptive mutation rates and crossover strategies tailored to genomic data variability are needed for robust optimization (Patel et al., 2023, BMC Bioinformatics).

In BioGA, I can try to address these gaps by:

• Implementing multi-objective optimization (e.g., NSGA-II).
• Adding parallel processing using OpenMP or RcppParallel.
• Incorporating gene network constraints for biologically relevant solutions.
• Supporting adaptive mutation and crossover strategies.
• Enhancing scalability for large genomic datasets.