Eukaryotic gene regulation at equilibrium, or non?

TL;DR

This paper discusses the limitations of equilibrium models in eukaryotic gene regulation and advocates for a normative, theory-driven approach to identify evolutionarily favored non-equilibrium mechanisms that better explain biological regulation.

Contribution

It proposes shifting from data-fitting to optimization-based modeling to focus on biologically relevant non-equilibrium gene regulation mechanisms in eukaryotes.

Findings

Equilibrium models underperform in eukaryotic gene regulation prediction.

Non-equilibrium mechanisms may be evolutionarily optimized for specific functions.

Simulation examples illustrate the potential of normative modeling approaches.

Abstract

Models of transcriptional regulation that assume equilibrium binding of transcription factors have been very successful at predicting gene expression from sequence in bacteria. However, analogous equilibrium models do not perform as well in eukaryotes, most likely due to the largely out-of-equilibrium nature of eukaryotic regulatory processes. These processes come with unavoidable energy expenditure at the molecular level to support precise, reliable, or fast gene expression responses that could correspond to evolutionarily optimized regulatory strategies. Unfortunately, the space of possible non-equilibrium mechanisms is vast and predominantly uninteresting. The key question is therefore how this space can be navigated efficiently, to focus on mechanisms and models that are biologically relevant. In this review, we advocate for the normative role of theory - theory that prescribes rather than just describes - in providing such a focus. Theory should expand its remit beyond inferring models from data (by fitting), towards identifying non-equilibrium gene regulatory schemes (by optimizing) that may have been evolutionarily selected due to their favorable functional characteristics which outperform regulation at equilibrium. This approach can help us navigate the expanding complexity of regulatory architectures, and refocus the questions about the observed biological mechanisms from how they work towards why they have evolved in the first place. We illustrate our reasoning by toy examples for which we provide simulation code.