Prediction of regime shifts under spatial indicators in gene transcription regulation systems

Abstract

Gene transcription regulation is a noisy process that can lead to regime shifts and these shifts are difficult to predict. In this letter, we consider spatiotemporal fluctuations in gene transcription regulation, and employ a spatially extended gene transcription regulation system by using theoretical analysis and numerical simulation. We find that the system produces a regime shift from a low (high) to a high (low) protein concentration state (i.e., the transform of the switch of gene transcription regulation), and we apply spatial early warning signals (e.g., rising spatial variance, spatial skewness, spatial kurtosis, and lag-1 autocorrelation) to study the regime shifts of protein concentration. The results show that the spatial early warning signals can be used as a robust indicator of predicting gene transcription regulation switch. Once the signal is detected, we will take measures to intervene. Our results may have a positive effect on improving early treatment interventions for complex human diseases.