Gapsplit

Gapsplit performs targeted random sampling of COBRA constraint-based models to achieve uniform coverage of convex and non-convex solution spaces for accurate model analysis.

Key Features:

• Targeted Sampling: Identifies and focuses sampling on under-sampled regions of the solution space to improve representation.
• Versatility Across Model Types: Handles both convex and non-convex constraint-based models used in metabolic networks.
• Uniform Coverage: Produces more uniform coverage of the solution space to support reliable analysis and interpretation.

Scientific Applications:

• Flux Balance Analysis: Enhances the accuracy of flux balance analysis by sampling a broader set of feasible flux distributions.
• Pathway and Regulatory Discovery: Aids identification of novel metabolic pathways or regulatory mechanisms by revealing under-sampled solution regions.
• Model Prediction and Simulation: Improves robustness of model predictions and simulations through more comprehensive sampling of feasible states.

Methodology:

Performs algorithmic identification and targeting of under-sampled regions using random sampling to achieve uniform coverage across convex and non-convex solution spaces of COBRA constraint-based models.