Research On Prediction Method Of The Interaction Between Microbe And Host

With the continuous development of biotechnology,people's cognition of life is also deepening.Especially with the emergence of the third generation gene sequencing technology,the massive gene data of various organisms has become an important basis for solving problems in medical health,agricultural breeding and environmental testing.However,severe acute respiratory syndrome(SARS),influenza A virus(H1N1),new coronavirus pneumonia(COVID-19)and other sudden illness outbreaks are still threatening the world's public health safety as the natural environment changes.How to effectively use bioinformation data in the prevention and control of major disease disasters is also one of the major challenges in the field of bioinformation.One of the solutions is to predict the interaction between bacteria and host through relevant data,so as to find the hidden intermediate host,effectively cut off the path of infection,and provide ideas for finding effective treatment methods.The prediction method of the interaction between bacteria and host has been improved and applied well.However,most of the current prediction methods are only for one kind of bacteria or host,and lack of prediction methods for the interaction between a variety of bacteria and a variety of hosts,which can not meet the current prediction needs.In order to solve these problems,we introduced the based on the Balanced KATZ measure and the Within and Between Score(BKATZWBS)for microbe-host interaction prediction.Firstly,the subject studied the Within and Between Score for Mi RNA-Disease Association prediction(WBSMDA)and based on KATZ measure to predict the Human Microbiota with non-infectious Diseases Associations(KATZHMDA).By using logistics transformation on the Gaussian interaction profile kernel similarity calculation result of microbe and host,and add balance factor to control the influence of Gaussian interaction profile kernel similarity of microbe and host on the final prediction result.Finally,combining the advantages of the two algorithms,the BKATZWBS is proposed.Next,we took the verified experiment of the prediction performance of BKATZWBS.The experimental results show that on the PHI-base data set,the relevant evidence of 17 pairs of microbe-host interactions in the top 20 prediction results of BKATZWBS can be found in the relevant literature.Afterwards,we adopted the leave-one-out cross-validation method on the PHI-base,HPIDB,and HMDAD data sets through the ROC curve and P-R curve,and the corresponding AUC value and AUPR value for comparing BKATZWBS,WBSMDA,KATZHMDA,Network Consistency Projection for Human Microbe-Disease Association prediction(NCPHMDA),Neighbor-and Graph based combined Recommendation model for Human Microbe–Disease Association prediction(NGRHMDA).The experimental results show that the AUC and AUPR values of BKATZWBS are higher than other comparison methods.The results show that the prediction results of BKATZWBS are ideal.Finally,the subject sorted out the bacteria-host interaction data and related data and designed the Bacteria-Host Protein Interaction Database(BHPID).The design purpose of BHPID is to facilitate the relevant scientific researchers to observe the bacteria-host interaction data more intuitively and reduce the time cost of data acquisition.Bacteria-host interactions data displayed in the table view,tree view and graph view.In addition,BHPID also provides search,submit and download functions.