Metabolic Network Sensitivity Analysis for Enzyme Yield Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing metabolic network design techniques struggle to accurately predict and propose enzyme modifications that enhance target substance yield, as flux balance analysis (FBA) lacks reaction kinetics and structural sensitivity analysis (SSA) is indefinite under strict conditions.
Innovation Solution
A metabolic network processing system that generates sensitivity matrices using a stoichiometric matrix, introduces a confidence level, and employs sequential Bayesian estimation to extract and propose modified enzymes with high confidence levels, verified through biological experiments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If flux balance analysis (FBA) is used to propose enzyme deletion and insertion, then the metabolic network design can be simplified, but enzyme modification cannot be proposed because reaction kinetics are not introduced
Solution Approach 1:
The patent combines FBA and SSA into a unified analysis system that can handle both enzyme deletion/insertion (FBA capability) and enzyme modification (SSA capability). The integration allows the system to propose all three types of enzyme changes within a single framework, resolving the limitation where FBA alone cannot propose modifications.
Solution Approach 2:
The integrated system achieves multi-functionality by enabling a single analysis platform to perform both FBA (for deletion/insertion proposals) and SSA (for modification proposals). This universal system can address all enzyme design needs without requiring separate analytical approaches.
2Reliability
If structural sensitivity analysis (SSA) is used to propose enzyme modification, then the prediction can be verified by biological experiments, but the analysis becomes indefinite under strict analysis conditions
Solution Approach 1:
The patent applies parameter changes by adjusting the strictness of SSA analysis conditions to an appropriate level. By optimizing the analysis conditions, the system achieves a balance where predictions remain verifiable through biological experiments while avoiding the indefiniteness that arises from overly strict conditions. This allows practical enzyme modification proposals with reasonable prediction accuracy.
3Ease of manufacture
If FBA is used for metabolic network design, then enzyme deletion and insertion can be proposed, but the proposal accuracy for enzyme modification is insufficient due to lack of reaction kinetics
Solution Approach 1:
The system merges FBA (easy to implement, good for deletion/insertion) with SSA (more accurate for modification proposals) into an integrated framework. This combination allows the system to maintain ease of use while improving the accuracy of enzyme modification proposals through the incorporation of structural sensitivity analysis.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure proposes, as an example, a metabolic network processing technique which is related to a metabolic network processing system for extracting and proposing a modified enzyme that achieves a target yield of a target substance in a metabolic network in order to constantly predict an increase or decrease in a yield of the target substance when a modification of the enzyme is proposed based on a structural sensitivity analysis. A control device executes (i) processing of generating N sensitivity matrices by using a stoichiometric matrix obtained by converting the metabolic network, a basis c of a kernel of the stoichiometric matrix, and a given probability distribution, (ii) processing of calculating, by checking a sign of each matrix element in the N sensitivity matrices and calculating a proportion of the sign in each matrix element, a confidence level of each matrix element, (iii) processing of extracting k matrix elements having high confidence levels among confidence levels for the number of matrix elements in the sensitivity matrix as modified enzymes, and (iv) processing of outputting the extracted k modified enzymes