Abstract:Saccharomyces cerevisiae, one of the most commonly used cell factories for industrial biotechnology, is widely employed for mass production of bio-based chemicals and value-added compounds. Due to complicated cellular metabolism and regulatory network of biological systems, genome evolution is generally required for engineering with complicated phenotypes, which are coordinated and regulated by multiple genes. To achieve rapid evolution of S. cerevisiae genome, this study employed Clustered Regularly Interspaced Short Palindromic Repeats Interference（CRISPRi） to regulate the expression of genes closely related to chromosome replication and maintenance, such as MSH2, TSA1, RAD27, and CLB5, known as mutator genes. By designing guide RNAs（gRNAs） with differential repression efficiency, four mutators: MSH2 mutator mainly for point mutations and small InDels（MM）, TAS1 mutator mainly for point mutations, small In Dels, and structural variants（TM）, RAD27 mutator mainly for small In Dels and structural variants（RM）, and CLB5 mutator mainly for structural variants and aneuploidy chromosomes（CM） were constructed to control genome mutation rates and types（e. g. point mutation, small In Dels, structural variants, and aneuploid chromosomes）. These mutators were used for continuous evolution of a series of industrially relevant phenotypes, such as isobutanol tolerance, xylose utilization, and β-carotene biosynthesis. Interestingly, TM was more efficient for evolving isobutanol tolerance, but TM and CM were preferred for evolving β-carotene overproduction, and all mutators were verified to have comparable performance for evolving xylose utilization with yeast strains. We also discovered that the effectiveness of mutators was dependent on the phenotype to be evolved. To address challenges in the evolution of phenotypes without pre-determined knowledge on mutation rates and types, a mixed mutator（MTRC） was constructed to rapidly evaluate the mutator-phenotype relationship. Finally, a combinatorial mutator（MTRC*2） were constructed to explore synergistic interactions among various mutators for the continuous genome evolution of S. cerevisiae. The established mutators can not only be used for constructing robust yeast cell factories, but also be further developed as a generally applicable genome evolution tool.

Context Matters: Adaptive Mutation for Grammars

Structured mutation inspired by evolutionary theory enriches population performance and diversity

Exploiting abstractions for grammar‐based learning of complex multi‐agent behaviours

An investigation into structured grammatical evolution initialisation

Mutator-driven Continuous Genome Evolution of Saccharomyces Cerevisiae

Towards Self-adaptive Mutation in Evolutionary Multi-Objective Algorithms

Model Approach to Grammatical Evolution: Theory and Case Study

Controlling the Mutation in Large Language Models for the Efficient Evolution of Algorithms

Modeling Grammatical Evolution by Automaton.

A Grammar Design Pattern for Arbitrary Program Synthesis Problems in Genetic Programming

Learning synchronous context-free grammars with multiple specialised non-terminals for hierarchical phrase-based translation

Effective Adaptive Mutation Rates for Program Synthesis

Learning of Structurally Unambiguous Probabilistic Grammars

A Rank based Adaptive Mutation in Genetic Algorithm

Improving NSGA-II with an Adaptive Mutation Operator

Resilient swarm behaviors via online evolution and behavior fusion

Study on Genetic Algorithm Based on Schema Mutation and Its Performance Analysis

Enhancing Genetic Improvement Mutations Using Large Language Models

Evolution of Scikit-Learn Pipelines with Dynamic Structured Grammatical Evolution

Automatic Design of Semantic Similarity Ensembles Using Grammatical Evolution

Geometric semantic genetic programming with normalized and standardized random programs