| Date: | 21.02.2017 |
| Time: | 14:30-18:00 (after the ITB meeting) |
| Room: | -02/03 Haus 4 (cellar1) |
| Material: | download all |
Tutorial on numerical methods in biology. This tutorial introduces numerical bifurcation analysis, diffusion approximations to Master equations and how to simulate them, on your computer. And you can bring your own models to analyse.
If possible bring your own computer, raspi, ubuntu phone, …. A recent installation of the following open sources packages is required:
pythonandipython.
In addition, the following python modules are needed
brian2,networkx,jupyter.
You can locally install this with, e.g.
pip install --user modulename.
It is also useful to visualise markov models of ion channels. This can be done with the module pygraphviz. pip would not install it on my machine due to some dependencies. But apt-get install graphviz libgraphviz-dev python-pygraphviz works.
And last but not least, download and install the following python modules:
They too can be installed with pip install --user modulename.tar.gz.
All numerical tools should be in place then2. Model equations to analyse are required. Here are some json-files with model definitions:
But you can bring your own. Ideally in the same format.
Also download the following ipython-notebooks.
The continuation software AUTO-07p needs to be compiled, so you need a fortran and a C compiler. For example sudo apt-get install gfortran gcc. Then, download AUTO here. Untar it, then use the ./configure script and make to compile it.
Documentations are found here: brian2 networkx.
Numerical analysis, a mathematical field, yes, but ipso facto with a deeply applied spirit. Its success depends on correct and efficient implementations. Most things we use in the tutorial (including the operating system3) were created by scientists around the planet in a joint efford. Open source software is only as good as the feedback from its user base. Who else, if not us should get involved?
Official git repos: brian2 networkx