Genetic Algorithm System for Diverse Molecular Variant Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for directed evolution of protein function are inefficient and prone to human bias, especially when dealing with large and diverse populations of molecular variants, making it difficult to identify variants with altered or improved properties against structurally different substrates or ligands.
Innovation Solution
A method is developed to create an optimized, diverse population of molecular variants by inputting desired mutations and setting optimization parameters such as population size, crossover probability, mutation rate, and fitness functions, using genetic algorithms to evolve and select variants that maximize diversity and information content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional directed evolution methods are used to generate protein variants, then variants can be obtained, but the process becomes inefficient and biased when dealing with large and diverse populations
Solution Approach 1:
The patent replaces manual, mechanical selection processes with an automated computer-based system that uses algorithms to generate, evaluate, and select molecular variants. The computer system automatically manages large populations of variants, performs computational assessments of desired properties, and identifies optimal candidates without human intervention, thereby improving efficiency while reducing the complexity burden of manual population management.
2Adaptability or versatility
If the population size of molecular variants is increased to improve diversity, then more variants are available for screening, but the difficulty of managing and evaluating the population increases
Solution Approach 1:
The patent introduces a computer-based evaluation system as an intermediary between the diverse molecular variant population and the researcher. This intermediary automatically assesses variants for desired properties using computational methods, managing the complexity of evaluating large populations while preserving and enabling the exploration of high diversity. The computer system acts as a mediator that handles the burden of population management, allowing researchers to work with diverse populations without being overwhelmed by evaluation complexity.
3Reliability
If manual selection of variants is performed, then human expertise can guide the process, but human bias limits the ability to objectively identify variants with novel activities
Solution Approach 1:
The patent replaces manual human selection with an automated computer-based selection system that objectively evaluates molecular variants for desired properties. This substitution eliminates human bias from the selection process, providing more reliable and objective identification of variants with novel or improved activities. The computer system can simultaneously evaluate large numbers of variants, maintaining high throughput while ensuring objective, bias-free selection based on defined criteria.
4Adaptability or versatility
If the number of mutations per variant is increased to enhance property improvement, then more diverse properties can be achieved, but the complexity of tracking and analyzing mutations increases
Solution Approach 1:
The patent introduces a computer-based system as an intermediary to track, manage, and analyze multiple mutations across diverse variants. This intermediary automatically records mutation data, tracks their combinations, and analyzes their effects on desired properties, thereby handling the complexity of mutation tracking while enabling the exploration of variants with multiple mutations and a broader range of altered properties.
Data Source
AI summary
The disclosure relates to a method of generating a diverse set of variants to screen improved and novel properties within the variant population, a system for creating the diverse set of variants, and the variant peptides.