Machine Learning Object Determination for Protein Evolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current directed evolution technologies for proteins are laborious and time-consuming, requiring extensive screening and recombination processes to achieve desired protein performance, resulting in high time and resource costs.
Innovation Solution
An object determining method using a computer device that acquires index prediction values, determines a mapping relationship between indices and object features, and selects target objects based on these relationships to efficiently identify proteins meeting specific performance criteria, leveraging machine learning and neural networks to streamline the process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional directed evolution methods are used to screen and recombine mutants, then proteins with desired functions can be obtained, but the process becomes laborious and time-consuming
Solution Approach 1:
The patent applies preliminary action by pre-training a machine learning model on existing protein structure and sequence data before the actual directed evolution process. This pre-trained model can predict mutant properties, allowing researchers to prioritize which mutants to screen experimentally, thereby reducing the overall time required while maintaining reliability in obtaining desired protein functions
Solution Approach 2:
The patent introduces a machine learning model as an intermediary between traditional screening methods and the final protein selection. This intermediary model predicts the effects of mutations and guides the screening process, reducing the number of experimental iterations needed while still achieving reliable identification of proteins with desired functions
2Reliability
If extensive screening and recombination processes are performed, then desired protein performance is achieved, but resource costs increase
Solution Approach 1:
The patent performs preliminary computational analysis using the trained machine learning model to predict which mutants are most likely to achieve desired performance. This allows researchers to focus experimental resources on a smaller, more promising subset of mutants, reducing overall resource consumption while maintaining the ability to achieve desired protein performance
Solution Approach 2:
The patent replaces extensive mechanical screening processes with computational predictions from the machine learning model. By substituting computational analysis for physical screening of all mutants, the method reduces resource consumption while still identifying proteins with the desired performance characteristics
Data Source
AI summary
Provided is an object determining method performed by a computer device, relating to the technical field of artificial intelligence. The method includes: acquiring index prediction values of objects in a first object set on a preset index respectively; determining, based on index experimental values and object features of the objects in the first object set on the preset index, a mapping relationship between the preset index and the object features; selecting, from the first object set, objects with the index prediction values satisfying index value screening conditions to obtain a second object set; and determining a target object meeting index requirements of the preset index from the second object set based on the mapping relationship.