GEM Engine: Simulation Landscapes and Optimization Benchmark Data for Genome-Scale Metabolic Design

GEM Engine: Publication Dataset This dataset supports the research manuscript "GEM Engine: A reusable landscape-guided framework for exploration and optimization of genome-scale metabolic design spaces" by Hamed Rastaghi and Samuel M. D. Oliveira. GEM Engine is a computational framework designed to support reusable, landscape-guided exploration and optimization of genome-scale metabolic models. The framework combines flux balance analysis (FBA), surrogate modeling, and multi-objective optimization to enable systematic investigation of metabolic design spaces while reducing redundant simulations across optimization tasks. Dataset Contents Results.zip Contains genome-scale metabolic simulation landscapes, multi-objective optimization benchmarks, and associated computational results supporting the study. Figures_and_Tables.zip Contains data and materials associated with the main and supplementary figures and tables, including supporting analysis outputs and generation scripts where available. Reproducibility and Software The archived results support comparisons of landscape-guided optimization strategies, including gsMOBO, NSGA-II, SPARC, GHOST, and TRAIL, as well as surrogate-quality-based algorithm selection. The GEM Engine source code is maintained separately on GitHub. The repository URL will be provided when the software is publicly released.