Research And Platform Construction Of Horizontal Gene Transfer Event Recognition Algorithm Based On Multithreading Technology

Horizontal gene transfer(HGT)events exist widely in nature,which helps recipient species to bypass mutation and recombination to get new genes and accelerate the process of genome innovation.Therefore,accurate identification of HGT events has become an important part of exploring the real evolutionary relationship between species.Since the locations of transferred genes tend to be preserved by pedigree,recognition of HGT events can be studied with the topological inconsistencies between gene trees and species trees.RIATA-HGT algorithm is currently one of the effective algorithms for identifying HGT events.However,this algorithm only calculates one gene tree data in the input gene tree set at a time,which requires a high time coste.In order to shorten its running time,this thesis adopts phylogenetic analysis method to improve RIATA-HGT algorithm based on multithreading technology.In this thesis,Kaikoura algorithm and Exponential Recursive(ER)algorithm are combined to propose a Multithreading-HGT(M-HGT)algorithm for HGT event recognition.The work of this thesis is as follows:For the input species tree and gene tree set,M-HGT algorithm allocates the identify HGT event task of species tree and each gene tree in the set to several threads through the thread pool.Each thread first verifies whether there is an element containing the input gene tree in the existing consensus phylogenetic network set through ER algorithm.And if not,the Kaikoura algorithm was used to calculate the maximum agreement subtree of the species tree and the gene tree.Then RIATA-HGT algorithm identify HGT events for the species tree and the gene tree.The M-HGT algorithm adds the HGT events recognized by each thread to the species tree in the form of edges to obtain the consensus phylogenetic network set.In this thesis,a user interaction platform is built,which integrates RIATA-HGT and M-HGT algorithms,so as to meet the needs of researchers to choose these two algorithms for horizontal gene transfer event recognition.In this thesis,experimental analyse of RIATA-HGT and M-HGT are carried out through the simulation data and the real biological data.The results show that the MHGT algorithm reduces the running time and improves efficiency while maintaining the same accuracy as the RIATA-HGT algorithm.