This repository can be used to simulate gene expression dynamics of self-repressing genes. It includes optimized code for deterministic and stochastic delay differential equations to describe gene expression oscillations, as well as code for parameter inference on these equations. Parameter inference can be conducted on single-cell time series data on protein expression through the use of a delay-adapted non-linear Kalman filter, or likelihood free through Approximate Bayesian Computation. Further model analysis methods hosted here include bifurcation and power spectrum calculations.

This repository contains supplementary material for the papers

• Quantitative single-cell live imaging links HES5 dynamics with cell-state and fate in murine neurogenesis, C. S. Manning, V. Biga, J. Boyd, J. Kursawe, B. Ymisson1, D. G. Spiller, C. M. Sanderson, T. Galla, M. Rattray, N. Papalopulu https://doi.org/10.1038/s41467-019-10734-8
• miR-9 mediated noise optimization of the her6 oscillator is needed for cell state progression in the Zebrafish hindbrain, X. Soto, V. Biga, J. Kursawe, R. Lea, P. Doostdar, N. Papalopulu https://doi.org/10.1101/608604
• Inferring kinetic parameters of oscillatory gene regulation from single cell time series data, J. Burton, C. S. Manning, M. Rattray, N. Papalopulu, J. Kursawe https://royalsocietypublishing.org/doi/10.1098/rsif.2021.0393

All computational results in these papers can be re-generated from here.

• /src/ contains the main Python module for this repository, hes5.py. It includes functions to simulate stochastic and deterministic model traces, to calculate summary statistics of expression and power spectra etc.
• /test/ contains the code to generate simulated data in the papers, and to make the paper figures. The code that was used in the papers by Manning et al. and Soto et al. is saved in the folders test_manning_et_al_2019 and test_soto_et_al_2019, respectively. Each of these paper-specific folders contains the following files
  • test_make_final_figures.py, which can be used to regenerate all figures in the paper from data saved in this repository and
  • test_make_analysis_for_paper.py, which can be used to generate all data needed in test_make_final_figures.py from scratch by re-running all simulations.
  • data/ is a folder that contains simulated data for each paper in the form of Python numpy arrays. In case of the paper by Soto et al., this folder is empty In case of the paper by Manning et al., users will need to download this folder from https://www.dropbox.com/sh/r7taosaoy821l1x/AABQHAWfX-uNGbqMRkAFw-r6a?dl=0. This is to overcome data storage limits imposed by Github.
  • All remaining files in these folders contain additional analysis that was conducted while working towards the paper.
• The file test_infrastructure.py contains functional tests for the most important functions in /src/hes5.py, and can be used to test whether the installation is working.

• nosetests
• matplotib
• pandas
• seaborn
• numba
• PyDDE (https://github.com/hensing/PyDDE)
• the ubuntu package python-tk (sudo apt install python-tk)

Please see the file troubleshooting.md for known installation errors.

We use the nosetests testing infrastructure to run all code in this project. Code can be run using the nostests command like this:

cd ./test
nosetests -s test_infrastructure.py

This will run all functions in test/test_infrastructure.py whose name starts with test_, and it is possible to toggle whether individual functions are run by renaming them, for example by replacing test_ with xest_. The same concept applies to all python modules in the /test folder. The file test_infrastructure contains a few functional tests for main functions in src/hes5.py, and the nosetests command above can be used to see whether the project is correctly set up.

Before running any tests in the subfolders test/test_soto_et_al_2019 and test/test_manning_et_al_2019, we advice to first download the folders /test/test_soto_et_al_2019/data and /test/test_manning_et_al_2019/data using the links https://www.dropbox.com/sh/5k8wofjk97uu5ux/AADhsd41JVzIL_KJTnnj5ysCa?dl=0 and https://www.dropbox.com/sh/r7taosaoy821l1x/AABQHAWfX-uNGbqMRkAFw-r6a?dl=0, respectively.

For any questions concerning this repository, please email Jochen Kursawe at jochen dot kursawe at st-andrews dot ac dot uk