Modelling microRNA regulation in time series data

Abstract

MicroRNA (mRNA) is a gene regulating factor that binds to messenger RNA (mRNA) and inhibits translation or causes degradation of the mRNA. Many methods are available to predict which mRNAs are targets for miRNA but most of them are derived from observing miRNA or mRNA behavior in conditions that are not natural to the miRNA/mRNA.

A method used to identify promoters active at cell differentiation was adapted to apply for miRNA regulation. Time series data for mRNA and computational target predictions were used to estimate miRNA activity in a multiple linear regression model and miRNA time series data was used to evaluate the results. Different filtering conditions for the miRNAs and mRNAs that were used in the model were also considered.

The estimated activity of miRNA was not in accordance with experimental data implying that the model was unsuccessful in predicting miRNA regulation. Excluding mRNAs with low expression did improve the results slightly but not enough for the unchanged model to be considered practical.