Bioretrosynthetic Pathway Prediction Using AND-OR Tree Expansion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing bioretrosynthetic methods are inefficient and labor-intensive, requiring complex parameter settings and numerous wet experiments, and often fail to ensure the actual availability of biosynthetic pathways due to low computational efficiency and lack of evaluation of metabolic reactions.
Innovation Solution
A bioretrosynthetic method and system based on an AND-OR tree and single-step reaction template prediction, where an OR node is selected as the product molecule, predicted using a pre-constructed single-step reaction template prediction model, and the AND-OR tree is expanded to generate synthetic pathways, reducing the need for complex parameter settings and improving computational efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional bioretrosynthesis methods are used to construct biosynthetic pathways by searching for similar reactions, then the method can find metabolic pathways, but the computational efficiency is low and it is not suitable for constructing pathways with long reaction steps
Solution Approach 1:
The patent segments the complex task of biosynthetic pathway construction into single-step reaction template prediction tasks. Instead of searching for complete multi-step pathways at once, the system predicts reaction templates one step at a time, significantly improving computational efficiency and enabling construction of pathways with longer reaction steps.
Solution Approach 2:
The patent pre-constructs a reaction template set containing metabolic reaction templates before the actual pathway construction. This preliminary preparation allows the system to quickly match and predict reaction templates during pathway construction without performing complex searches in real-time, thereby improving computational efficiency.
2Ease of manufacture
If existing bioretrosynthetic methods are used, then biosynthetic pathways can be generated, but complex parameter settings such as thermodynamics, cofactors, and enzyme performance are required, increasing domain knowledge requirements
Solution Approach 1:
The patent extracts and removes the requirement for complex parameter settings (thermodynamics, cofactors, enzyme performance) from the biosynthetic pathway construction process. The system achieves pathway construction through single-step reaction template prediction without needing these complex parameters, thereby simplifying the process and reducing domain knowledge requirements.
Solution Approach 2:
The patent uses simple reaction templates as disposable building blocks for constructing biosynthetic pathways. Instead of requiring complex and expensive parameter calculations, the system uses straightforward reaction template matching that is computationally inexpensive and easy to implement, making the process more accessible to biologists.
3Reliability
If existing bioretrosynthetic methods are used to search for metabolic reactions, then reaction pathways can be found, but the correctness of metabolic reactions obtained by each step is not evaluated, making it difficult to ensure actual availability
Solution Approach 1:
The patent introduces a feedback mechanism that evaluates the correctness of metabolic reactions at each step of pathway construction. The system assesses whether predicted reaction templates are biologically valid and can actually be implemented, providing feedback that ensures the reliability and actual availability of the constructed biosynthetic pathways while maintaining construction efficiency.
Data Source
AI summary
The present disclosure provides a bioretrosynthetic method and system based on an AND-OR tree and single-step reaction template prediction. The method decomposes generation of a retrosynthetic pathway of a target molecule into multiple steps of the single-step reaction template prediction when conducting retrosynthesis on the target molecule. During the single-step reaction template prediction, a substrate molecule of the reaction template predicted in a previous step is used as a product molecule of a current reaction template to be predicted. Molecular characteristics of the product molecule are subjected to custom calculation using a SMILES sequence of the product molecule as an input, to be compatible with various single-step reaction template prediction models. The reaction template is determined according to a prediction result of the model, and extended based on a structure of the AND-OR tree. As a result, a potential synthetic pathway is found for the target molecule.