# DInference-based assessment of identifiability in nonlinear biological models

## On this page

This tutorial describes how to download and execute the inferential analyses as
described in “Inference-based assessment of identifiability in nonlinear biological models”,
submitted to *Journal of the Royal Society Interface* in 2017.

### Installation

This project requires the Functional Curation add-on to Chaste in order to run, which in turn requires the Chaste source tree to be installed. Instructions for this installation can be found for a variety of operating systems under InstallGuides/.

Extra install commands needed as well as the Ubuntu chaste-dependencies package are (on 14.04 at least):

Afterwards, obtain the latest version of all the code from the Chaste repositories using:

### Usage

Source code for the parameter fitting algorithms and plotting routines is contained in the `src`

folder.
Python scripts for performing inference on all model problems can be found in `tests`

.

An annotated CellML file describing the Hodgkin-Huxley model problem can be found in
the top-level project directory as `hodgkin_huxley.cellml`

, while functional curation
protocol files that describe the simulation and recording thereof are contained within
`tests/protocols`

.

A description of important files and their contents follows below.

Files relevant for implementation of the samplers:

`src/python/modeling/fitting/MCMC.py`

contains the implementation of the adaptice covariance MCMC sampler.`src/python/modeling/fitting/approximatebayes.py`

contains the implementation of the Del Moral ABC-SMC sampler.`src/python/modeling/fitting/objective.py`

contains the implementation of the objective functions used for inference

Files relevant for inference on the logistic model:

`src/python/modeling/simulation/LogisticGrowthExperiment.py`

defines the logistic model.`test/Logistic.py`

contains all fitting experiments associated with this model as described in the paper.

Files relevant for inference on the Hodgkin-Huxley model:

`hodgkin-huxley.cellml`

in the top-level directory contains a CellML description of the equations governing the Hodgkin-Huxley model.`test/HHVoltage.py`

contains all fitting experiments involving voltage data described in Section 4.1 of the paper.`test/HHSumStats.py`

contains all fitting experiments involving summary statistic data described in Section 4.2 of the paper.

To generate all described posterior estimates for the logistic model as described in the paper, go to the top-level Chaste directory and use:

To generate all described posterior estimates for the Hodgkin-Huxley model as described in the paper, use:

To see verbose output on the progress of the fitting algorithms, add the flag `no_store_results=1`

to the `scons`

commands above.
Note however that this will prevent storing a copy of the output on disk.