Research On Second Identification Algorithm Of Genome Island Based On Mobile Elements

Genomic islands(GIs),as an important vector in the process of horizontal gene transfer,can be transferred into the new host genome by transformation,transduction and so on.Compared with host genomes,genome islands have different characteristic information and often carry functional genes that have related to biological evolution and adaptation to organisms.With the transfer of horizontal genes,it may promote communication between different species,which is of great significance for studying microbial evolution and functional analysis.The genomic island prediction methods based on sequence composition and the genomic island prediction method based on comparative genomics,ignored the role of Mobile elements.In this dessertation,based on existed prediction methods for genomic island,we proposed a secondary recognition strategy for genomic islands based on Mobile elements,and analyzed association between the recognition strategy and the prediction method.Then,we proposed the optimal recognition strategy.The specific tasks are as follows:1.We summarized the database of genomic island and the database of mobile elements,systematically collected the data of five kinds of Mobile elements,such as inserted sequence,ICE,MGE,integron,phage.2.We summarized several prediction methods for genomic island in detail,which are divided into two groups.The results of each prediction method are obtained.We discussed the prediction method of advantages and disadvantages,which provides theoretical basis for following optimization.We summarized the database of genomic island and the database of mobile elements,systematically collected the data of five kinds of Mobile elements,such as inserted sequence,ICE,MGE,integron,phage.3.We proposed a prediction methods for genomic islands based on sequence composition and mobile elements.Based on the prediction methods based on sequence composition,we designed filter and increase recognition strategy,and realized the secondary recognition strategy for genomic islands by mobile elements.The accuracy of secondary recognition algorithm improved more than 15% in S.enterica serovar typhi CT18 genome except Alien Hunter.InGI/non-GI dataset,The recall of secondary recognition algorithm improved more than 20% and the accuracy improved more than 82%.The TPR of proposed method improved more than10% in L-data.Adding GIs containing ICE and phage through sliding windows can increase efficiency of method based on sequence composition.4.We proposed a prediction method for genomic islands based on comparative genomics and mobile elements.First,we calculated component vector through sling windows and determined reference genome according to distance between species.We then selected the conserved regions using Mauve.Finally,we constructed the secondary recognition algorithm of genomic island by adding mobile element assisted optimization.Experimental results show that the result of proposed method performed better than method based on comparative genomics.The result indicate that the efficiency of the recognition algorithm based on comparative genomics could be improved by filtering out the genomic islands that do not contained ICE and selecting the candidate genomic islands that contained ICE and phage.