Microbiota Mix Prediction Using Evolutionary Ratio Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting and producing microbiota mixes are inefficient, often empirical, consume rare materials, and require substantial time due to lengthy analysis and co-cultivation processes, while existing prediction models are not invertible and unsatisfactory.
Innovation Solution
A computer-aided method using linear and non-linear prediction models to simulate microbiota mix compositions, involving matrix calculations and evolutionary algorithms to determine initial samples and mixing ratios, allowing instant and accurate simulation without material consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If test-based approach with random mixing and sequencing is used, then final mix profiles can be obtained, but it consumes rare materials and takes several weeks to be completed
Solution Approach 1:
The patent applies preliminary action by using prediction models to forecast mix profiles before actual mixing and sequencing occurs. The model predicts the outcome of mixing based on input sample profiles, allowing virtual testing of mix compositions without physically preparing and sequencing every possible mix, thus saving time and rare materials while maintaining profile accuracy.
Solution Approach 2:
The patent uses computational models to create virtual copies of mix profiles instead of physically producing and sequencing actual mixes. The prediction model generates synthetic profile data that represents the expected composition of mixed microbiota samples, eliminating the need for physical sample preparation and sequencing while maintaining measurement precision.
2Adaptability or versatility
If expansion through co-cultivation in multiple bioreactors is performed, then microbial diversity is maintained, but it requires substantial co-cultivation time lengthening test durations
Solution Approach 1:
The patent applies preliminary action by predicting the outcomes of co-cultivation processes before they are executed. The prediction model takes input profiles of microbiota samples and predicts the resulting mix profiles after co-cultivation, allowing virtual assessment of diversity and composition without actually performing the time-consuming co-cultivation experiments in multiple bioreactors.
Solution Approach 2:
The patent replaces the physical mechanical process of co-cultivation in bioreactors with a computational prediction system. Instead of physically cultivating microorganisms in controlled environments for extended periods, the system uses algorithmic models to simulate and predict the outcomes of such cultivation processes, maintaining diversity assessment capability while eliminating the time requirement.
3Measurement precision
If known prediction models are used, then mix composition can be predicted, but the models are not invertible and are not satisfactory
Solution Approach 1:
The patent applies inversion by developing prediction models that can operate in reverse - instead of only predicting forward from input samples to mix composition, the model can be inverted to determine what input samples are needed to achieve a target mix composition. This invertibility is achieved through mathematical formulations that allow bidirectional computation, enhancing both prediction accuracy and adaptability for different application scenarios.
Data Source
AI summary
An evolutionary algorithm is used to determine parameters of a production process of a complex microorganism community, CMC, product given a target profile for the CMC product. CMC mixing operation and co-cultivation operation of a CM C product are modelled, using learnt matrix-based models. The evolutionary algorithm iteratively modify candidates representing the parameters, including a set of complex microorganism community samples in the initial sample collection and mixing ratios for one or more mixing operations in the production process. The determined set of samples and associated mixing ratios are then used to control actual picking and processing of complex microorganism community samples according to the mix production process, to obtain a CM C product as close as possible, in terms of profiling features, to the target profile.