Research On Identification Algorithm Of Essential Proteins Based On Gene Ontology And Topology Structure

Essent ia l proteins are necessary in the surviva l and reproduct ion of organis ms,the ident ificatio n of t he m can he lp us to understand the minimum require ments o f the ce llular life.Wit h the improve ment of protein-protein interact ion data,the ident ificat ion o f essent ia l proteins based on network topology have attracted widespread attent ion,but due to the fact that the data of protein network is not comp lete and its high fa lse posit ive,the recognit io n rate of essent ia l proteins have been affected.Cons idering t he biolo gical funct ions and properties of the nodes in network,we treat d iffere nt ly to a ll the edges in t he network.The paper uses the informat io n of Gene Onto logy to calculate the funct iona l similar ity between two endpoints proteins based on the Ge ne Ontolo gy ter ms,and it is used to measure the reliability of the protein interaction.The paper ident ifies essentia l prote ins while giving different weights to each edge of the network and comb ining wit h topologica l features of the network.In this paper,two works are inc luded :Because of t he fact that some fa lse posit ive informat io n exist in protein-protein interaction net works and essentia l protein's neighborhood have copolymer izat ion class feat ures.Comb ining t he Gene onto logy informat io n and t he edge cluster ing coeffic ient,we present a new a lgor it hm,the EGC a lgor it hm(Edge cluster ing coefficient and Gene ontology information's Combination).The experimental results show that EGC has a higher ident ificat ion rate of essent ia l proteins than other met hods in two yeast protein data sets(DIP and MIPS)and it also can ident ify the essent ia l proteins ignored by other methods.According to t he fact that essent ia l proteins appearing a ggre gat ion in the sa me or similar biolo gical funct iona l protein comp lexes.Comb ining the Gene ontology informat io n and comp lex infor mat ion,we present a new algorit hm,the CCG algor it hm(Comb ining Comp lex centra lit y and Gene ontology infor mat ion).The experimenta l results show that CCG can ident ify more essent ia l proteins and perfor ms better on six statist ica l ind icators than other met hods in three yeast protein data sets(DIP?MIPS and BioGRID)and two protein comp lex sets(CM270 and CM408).