Case study / scientific software
A modular, FAIR-compliant MATLAB framework for ODE biochemical pathway modeling, simulation, parameter estimation, global sensitivity analysis, and profile likelihood analysis. Thesis: A workflow for developing biochemical pathway models using ordinary differential equations.
Problem
ODE-based biological models can be hard to build, parameterize, analyze, and share when each lab relies on bespoke scripts. The goal was a workflow that supports initialization, simulation, parameter estimation, sensitivity analysis, and profile likelihood analysis while following FAIR principles.
Contribution
I architected and built the framework, implemented algorithms from the scientific literature, engineered interfaces between MATLAB solvers and COPASI, created documentation, and managed Git/GitHub collaboration across institutions.
Reusable ODE modeling pipeline
The workflow connects model initialization, simulation, parameter estimation, local and global sensitivity analysis, and profile likelihood analysis into a reproducible scientific software process.
Implementation details
Structured model initialization and data handling for biochemical pathway models, with attention to standard formats like SBML and SBtab.
Methods for parameter estimation and model calibration using MATLAB and scientific optimization tooling.
Local and global sensitivity analysis plus profile likelihood analysis to understand model behavior and parameter identifiability.
Validation
Ligand-receptor interactions and ligand-receptor-effector responses validated as small reference models for the workflow.
Translated complex CaMKII autophosphorylation dynamics into ODE form for medium-scale validation.
Pushed the framework into larger parameter spaces: intensive parameter estimation and PLA on an Intel i9-10980XE workstation. Resulted in a peer-reviewed Neuroinformatics paper.
Workflow
The framework follows FAIR principles for findable, accessible, interoperable, and reusable research software, integrates standard formats, and uses Git workflows for cross-institutional collaboration.
Related work
Browse all projects or review the CV for the broader research and publication context.
Case study · scientific software
A modular, FAIR-compliant MATLAB framework for ODE biochemical pathway modeling, simulation, parameter estimation, global sensitivity analysis, and profile likelihood analysis. Thesis: A workflow for developing biochemical pathway models using ordinary differential equations.
Setup
Model + data
Simulation
MATLAB solvers
Estimation
Optimization
Analysis
Sensitivity / PLA
Standards
SBML / SBtab
Output
Validated results
Problem
ODE-based biological models can be hard to build, parameterize, analyze, and share when each lab relies on bespoke scripts. The goal was a workflow that supports initialization, simulation, parameter estimation, sensitivity analysis, and profile likelihood analysis while following FAIR principles.
Contribution
I architected and built the framework, implemented algorithms from the scientific literature, engineered interfaces between MATLAB solvers and COPASI, created documentation, and managed Git/GitHub collaboration across institutions.
Implementation details
Structured model initialization and data handling for biochemical pathway models, with attention to standard formats like SBML and SBtab.
Methods for parameter estimation and model calibration using MATLAB and scientific optimization tooling.
Local and global sensitivity analysis plus profile likelihood analysis to understand model behavior and parameter identifiability.
Validation
Ligand-receptor interactions and ligand-receptor-effector responses validated as small reference models for the workflow.
Translated complex CaMKII autophosphorylation dynamics into ODE form for medium-scale validation.
Pushed the framework into larger parameter spaces: intensive parameter estimation and PLA on an Intel i9-10980XE workstation. Resulted in a peer-reviewed Neuroinformatics paper.
Workflow
The framework follows FAIR principles for findable, accessible, interoperable, and reusable research software, integrates standard formats, and uses Git workflows for cross-institutional collaboration.
Stack
Related work
Browse all projects or review the CV for the broader research and publication context.