Abstract
There are two kinds of interactions between genes and metabolites. One is regular interactions, and the other is conditional ones. In this study, we extract the characteristic interactions that depend on specific stress conditions.
Based on a state-space model that evaluates the transitions of the time-dependent interactions from a probabilistic view point, the time points at which interactions are changed are detected. Then, genes and metabolites are extracted that relates to the transition time points. Conclusively, conditional interactions are compared among the stress conditions. We suggest a statistical approach to analyze these processes systematically.
We apply the proposed method to time series data of Arabidopsis thaliana stress responses and extract genes and metabolites that express functions that depend on specific stress conditions.
By means of the extraction of genes and metabolites related to some specific stress responses, the regulation for the adaptation to various environmental conditions will be able to be verified.
