Model Identifiability, Sensitivity & Uncertainty — Hands-on

Learn how to evaluate the reliability of systems biology models via identifiability, sensitivity and uncertainty analysis. This module covers structural vs practical identifiability, local and global sensitivity methods, uncertainty propagation and how to communicate model confidence in research and translational settings.

Model Identifiability, Sensitivity & Uncertainty

Help Desk · WhatsApp

Session Index

Session 1 — Identifiability Concepts & Diagnostics Session 2 — Local Sensitivity Analysis Session 3 — Global Sensitivity & Uncertainty Propagation Session 4 — Mini Capstone: Reliability Assessment of a Model

Session 1

Fee: Rs 8800

Identifiability Concepts & Diagnostics

What identifiability means for biological models

parameter vs prediction focus structural vs practical identifiability role of data richness

Simple examples of identifiable and non identifiable models

lumped vs separate parameters observability and measurements design intuition from examples

Basic diagnostics and red flags in practice

flat cost surfaces non unique parameter sets correlated parameters

Session 2

Fee: Rs 11800

Local Sensitivity Analysis

Local sensitivity coefficients and interpretation

dimensionless sensitivities time dependent profiles ranking influential parameters

Computation via finite differences and built in tools

forward difference basics step size considerations automatic differentiation overview

Using local sensitivity for experiment design ideas

which parameters need better data candidate readouts prioritizing perturbations

Session 3

Fee: Rs 14800

Global Sensitivity & Uncertainty Propagation

Global sensitivity concepts and sampling strategies

parameter ranges and priors Latin hypercube sampling Monte Carlo basics

Sobol style indices and variance based methods

first order and total effects interaction insights interpretation for modelers

Uncertainty propagation to model outputs

prediction bands and envelopes quantiles and credible intervals communicating uncertainty clearly

Session 4

Fee: Rs 18800

Mini Capstone: Reliability Assessment of a Model

Select an ODE or FBA linked model and define reliability questions

Theory plus guided practical

Perform sensitivity and basic uncertainty analysis on key outputs

local sensitivity profiles simple global screening prediction bands for outputs

Deliverables: notebook, summary tables and reliability note

Python or R notebook CSV of sensitivity metrics PDF or HTML reliability summary

PDF

Model Identifiability, Sensitivity & Uncertainty Training | Reliable Systems Biology Models

Model Identifiability, Sensitivity & Uncertainty — Hands-on