Phage Tail Module Recognition Based On Hidden Markov Model

Bacteria play an important role in nature,the presence of bacteria in human organs such as the gut and skin helps us perform various physiological and biochemical functions that we can't do on our own.On the other hand,a class of pathogenic bacteria in bacteria can also cause our diseases.Even if the emergence of antibiotics against the disease-causing ability of these pathogenic bacteria,the effectiveness of antibiotics will be greatly reduced with the emergence of bacterial resistance.Therefore,a kind of natural anti-bacterial organism,bacteriophage,gradually appeared in people's vision.Through the modification of bacteriophage,we can make the bacteriophage adapt to different bacterial environment,so as to achieve the expected function.In this paper,the bioinformatics method was used to construct a model,and the location of the tail module of phage was predicted by the model,so as to provide the tail information needed for phage modification.Modified phages can specifically recognize a particular pathogen to treat a disease,or transfer specific genetic material into the bacteria.Hidden Markov model is a statistical probability model,which can be used to represent an observation sequence.At present,it has become a widely used tool model in time series data modeling In this paper,the three main parameters of the Hidden Markov Model of 6287 tailed phages were constructed: transfer probability matrix emission probability matrix initial probability matrix.At the same time,the tail protein domain database and non-tail protein domain database characterized by PFAM were also constructed The Viterbi algorithm is used to predict the tail protein of the hage.This algorithm is the main algorithm to solve the decoding problem in the hidden Markov model.The core idea is dynamic programming and recursionThen DBSCAN density clustering was performed on 961 verification set phages.The values of the two parameters of density clustering were determined through preliminary experiments in the early stage.Modularity of tail protein was determined through analysis and gene was divided into several large families,and obtained some rules through the analysis of the protein arrangement in the tail module.It was found that the arrangement of tail fibers in phage was relatively more concentrated and continuous,and the starting protein and stopping protein of the whole tail module were usually composed of the tail substrate or tail fiber.The standalone version of the whole tool is built and good visualization results are provided.The python program combining multiple steps is also concentrated in a script file,which is convenient for users to use multi-phage batch prediction provided in the tool and can also provide more results and more accurate rules for researchers to use.