Inferring Device for Automated Structural Formula Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating structural formulas using machine learning algorithms face challenges in automatically producing compounds with desirable chemical properties, requiring extensive iterations and time to achieve favorable results.
Innovation Solution
An inferring device that acquires latent variables, generates structural formulas, and calculates scores, repeating the process to improve the structural formula's score, utilizing models like Junction Tree VAE and AAE, and employing docking simulations to evaluate chemical properties and optimize the model for better compound generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine learning algorithms are used to generate structural formulas, then structural formula generation becomes possible, but it takes quite a long time to automatically generate structural formulas with preferable chemical properties
Solution Approach 1:
The patent implements a feedback mechanism where the score calculated from chemical properties is used to iteratively improve the latent variable and regenerate structural formulas. The system continuously refines the generated formulas by using the score as feedback to guide the optimization process, enabling automatic generation of high-quality structural formulas without manual intervention
Solution Approach 2:
The patent performs preliminary actions by pre-calculating chemical properties and scores for generated structural formulas before final selection. The system evaluates multiple candidates in advance using docking simulations and other chemical property assessments, allowing it to quickly identify and select the most preferable structural formulas without extensive iterative regeneration
2Manufacturing precision
If extensive iterations are performed to generate structural formulas with preferable chemical properties, then the quality of generated compounds improves, but the time and computational resources required increase significantly
Solution Approach 1:
The patent changes parameters by optimizing the balance between the number of iterations and the quality threshold. The system dynamically adjusts generation parameters based on the score improvement rate, stopping iterations when the marginal gain in quality no longer justifies the computational cost, thus maintaining high quality while improving productivity
Solution Approach 2:
The patent applies partial action by performing a limited number of iterations focused on the most promising candidates rather than exhaustively searching all possibilities. The system generates an initial set of structural formulas, evaluates them, and only performs additional iterations on those showing potential, thereby achieving high quality results with reduced computational resources and time
Data Source
AI summary
An inferring device includes one or more memories and one or more processors. The one or more processors are configured to acquire a latent variable; generate a structural formula by inputting the latent variable in a first model; and calculate a score with respect to the structural formula. The one or more processors execute processing of the acquisition of the latent variable, the generation of the structural formula, and the calculation of the score, at least two times or more, to generate the structural formula indicating the score higher than that of the structural formula generated at the execution of the first time.