2024-07-10 –, Method (1.5)
In this poster, a large-scale mathematical model of the JAK-STAT pathway was implemented in Julia to explore IFN-I ligand discrimination. Model fitting, often time-consuming in large-scale models, was performed to different temporal IFN-I inputs. Julia’s efficiency significantly accelerated this process by expediting sensitivity analysis, parameter estimation and identifiability analysis. The resulting model facilitates investigation into the mechanisms underlying ligand discrimination.
Type I interferons (IFN-I) play a crucial role in the cell's antiviral, antiproliferative, and immunomodulatory functions. Although multiple IFN-I subtypes activate the JAK-STAT pathway through the same receptor, they induce distinct downstream activation patterns. The mechanisms responsible for this ligand discrimination remain largely unclear. Hence, we employed the Julia Language to adapt the ODE model of the JAK-STAT pathway developed by Kok et al. to investigate IFN-I discrimination, focusing on IFNα and IFNβ. This model, consisting of 41 state variables and 73 parameters, was directly imported into Julia from SBML.
Various temporal inputs of IFN-I, including sustained, single-pulse and double-pulse stimuli, were used to calibrate the model. Parameter estimation was performed to improve model fitting to this temporal data, both local and global sensitivity analysis were employed to acquire candidate parameters. Next, the optimal parameter set was determined based on the Akaike Information Criteria. Finally, Profile Likelihood Analysis was employed to assess the identifiability of these parameters. The utilization of the Julia language significantly expedited these computational processes. The resulting model provides a tool for exploring the mechanisms underlying the ligand discrimination of IFN-I.
