Genome optimization and evolution simulation using genetic algorithms and GA-TRMR

作     者：Mortazavi, Ponehsadat 

作者单位：University of Colorado 

学位级别：master

导师姓名：Clauset,Aaron,eadvisorDowell-Deen,Robin Deanneecommittee memberGill,Ryan T.ecommittee member

授予年度：2013年

主      题：Evolution simulation Gene network Genetic Algorithms Genome optimization Synthetic fitness landscape Trackable Multiplex Recombineering Artificial Intelligence and Robotics Bioinformatics Computer Sciences Genetics 

摘      要：There are various desirable traits in organisms that humans wish to improve. To change a trait, the genetic background affecting that trait must be manipulated. In this thesis, genetic algorithms (GA) is suggested as a search strategy to efficiently sample from the combinatorial sequence space to identify the optimum genome sequence. Also, we proposed a restricted form of GA that approximates the search of a Trackable Multiplex Recombineering (TRMR) based approach, and is named GA-TRMR. The performance of GA and GA-TRMR strategies are simulated on various synthetic fitness landscapes through generations. The landscapes are constructed using fitness functions modeling for various degrees of epistasis. Both algorithms demonstrated viability for finding the optimum sequence. GA-TRMR specifically, when is equipped with a form of local sampling algorithm along with a selite set, turns into a powerful search algorithm which could find the optimum genome sequence in interaction with TRMR. Advisors/Committee Members: Aaron Clauset, Robin Deanne Dowell-Deen, Ryan T. Gill