Research On Constructing Paleontological Phylogenetic Tree With Missing Data And Inapplicable Data

A paleontological phylogenetic tree is the important contents in life science,which is used to infer the evolutionary relationships of taxon.For the early paleontology,constructing a paleontological phylogenetic tree can only rely on the fossil-observable morphological data.Paleontology morphological data,however,often own a large number of missing and inapplicable data,which makes difficult to construct a stable paleontological phylogenetic tree.To solve this problem,this paper proceeds from the characteristics of paleontological morphological data characteristics,and deals with the missing data and inapplicable data of paleontological morphological data rationally,and then infers phylogenetic relationships based on morphological characteristics to construct a paleontological tree.The specific content of this paper is as follows:(1)According to the features of paleontological morphological data,in this paper we proposes a hierarchical analysis method for the characteristics of paleontological morphological data to establish a hierarchical feature map of paleontological.Combining with hierarchical feature maps,the missing data of paleontology morphological data are divided into master feature missing data and slave feature missing data,and we also point out the the essential difference between missing data and inapplicable data of paleontological morphology that the existence of missing data is meaningful;but,inapplicable data is meaningless.(2)In this paper we proposes a missing data filling algorithm based on hierarchical relationship and cyclic iterative prediction model,combining with the hierarchical analysis of paleontological morphological data features,missing data missing mechanisms and missing patterns,to fill in the master feature missing data and slave feature missing data,because of the paleontology morphological data often own a large number of missing and inapplicable data,which makes difficult to construct a stable paleontological phylogenetic tree.We verify the validity of the missing data filling algorithm of paleontology morphological data,by missing the complete data randomly.(3)We propose a algorithm of constructing paleontological phylogenetic tree based on hierarchical features and parsimonious principles,combines with the feature hierarchization and phylogenetic analysis.Firstly,analysing the hierarchical feature of paleontology morphological and preprocessing the feature matrix of morphological data,a complete morphological data feature matrix of paleontology is obtained;Then,the characteristic is polaritied,through the comparison of outgroups to determine the ancestor feature and derive feature.Finally,choosing the useful features based on hierarchical features and parsimony principles to infer the evolutionary relationships of taxon,we construct a paleontological phylogenetic tree.Through the measurement of the constructed paleontological phylogenetic tree,we provide more references for palaeontologists on studying the evolutionary relationship of taxon.