Two methods for solving models of gene expression show promise
Two methods for solving models of gene expression show promise lead image
Gene expression is the randomly determined process in which information in a section of DNA is transformed into a protein molecule or RNA. Unfortunately, modeling gene expression is complex due to its stochastic nature, and even simplified models based on approximate solutions often fail to capture crucial observations.
Ham et al. demonstrate two quick and computationally efficient approaches for solving certain classes of models of gene transcription. One approach is based on modeling approximate solutions of gene expression models and the other is based on exact solutions.
“The analytical models can be straightforwardly fitted to data arising from the appropriate gene systems and used to progress our understanding of these fundamental processes,” said author Lucy Ham.
The authors found complicated multistate models can be broken down into simpler processes which can be individually solved. This allows researchers to find solutions for a wide range of multistate models. The authors also present a recurrence method to approximate the solution of a gene expression model based on a power series expansion of the stationary distribution.
Compared to the stochastic simulation algorithm, the recurrence method was found to be several orders of magnitude more efficient and compare more favorably to the well-known finite state projection algorithm in terms of computational cost in certain scenarios. The authors also identified a way to assess the importance of extrinsic noise in experimental data.
“In the future, we’re interested in researching questions regarding the role of extrinsic noise on gene systems, such as ‘How much variation is there in gene activity, both across a cell population, or over time within an individual cell?’” said Ham.
Source: “Exactly solvable models of stochastic gene expression,” by Lucy Ham, David Schnoerr, Rowan Brackston and Michael P. H. Stumpf, Journal of Chemical Physics (2020). The article can be accessed at https://doi.org/10.1063/1.5143540 .
