On the Origins and Rarity of Locally but Not Globally Identifiable Parameters in Biological Modeling

Contributions to glucagon and pramlintide pharmacokinetics and pharmacodynamics modeling for multi-hormone artificial pancreas systems

“…If the number of those parameter sets is infinite, the system becomes structurally unidentifiable. Checking local identifiability is usually easier than assessing global identifiability, and a recent work by Barreiro and Villaverde (2023) posed whether ensuring SLI would suffice since it is often the case that an SLI parameter is SGI too. The work analyzes 102 biological models from the literature to analyze their structural identifiability, and concluded that in 92.4% of the cases, a SLI parameter was also SGI.…”