Method for constructing phenotype ontology based on time series data

Abstract

The invention discloses a method for constructing a phenotype ontology based on time series data. The method is used for solving the technical problem that the existing phenotype ontology constructionmethods have poor practicability. The method comprises the following steps: instantaneous phenotypes are automatically identified and captured through the interaction relationship between genes, therelationship between the instantaneous phenotypes is captured to construct a directed acyclic graph, and the phenotype ontology is finally obtained. The method is developed around the plant phenotypebody, and the time series phenotype ontology of the plant is constructed through the relationship among the genes, phenotypes and environment. The method is based on a graph theory to mine potential patterns in the time series phenotypic data of the plant, and has the advantages of easiness in algorithm realization, low time complexity, and completion of the construction of the ontology construction within limited time complexity and space complexity. The phenotype ontology constructed by the method has a complete topology structure, and is the directed acyclic graph annotated with biologicalinformation. The relationship between nodes is well preserved, the biological information and the biological meaning of the nodes are also well annotated, and the practicality is good.

Description

technical field [0001] The invention relates to a method for constructing a phenotype ontology, in particular to a method for constructing a phenotype ontology based on time series data. Background technique [0002] The document "Inter-Functional Analysis of High-throughput Phenotype Data by Nonparametric Clustering and its Application to Photosynthesis. Bioinformatics, 2015: btv515" discloses a new clustering method NPM (Non-parametric Model) based on non-parametric modeling, which uses Clustering techniques to analyze phenotypic data. For mutants with deliberate gene knockouts, the NPM algorithm initially selects anchor points at random, and then combines Gaussian kernel functions to define a suitable kernel function for each anchor point. Then use the EM algorithm to iterate continuously to complete the clustering. Among them, the "E step" uses a posterior probability to continuously estimate the probability that the sample belongs to the current class, and uses this p...

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the shortcomings of poor practicability of existing phenotypic ontology construction methods, the present invention provides a time-series data-based phenotypic ontology construction method

Images