Overview: MyGenes is an environment dependant evolution of digital organisms simulator. This script and GUI were developed for my MSc dissertation and shared to assist in new investigations on this topic.
Mygenes by Daniel VJ is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
To use My Genes on your operating system you'll need to have Python 3 installed. Make sure NumPy and SciPy. If you have no idea how to install these, I suggest downloading the Python 3 distribution by Anaconda.
After the installation, make sure Python is installed correctly on your system by opening the Terminal (or cmd.exe on Windows), type "python" or "python3" and then hit Enter. If it shows something like the image below, it means you're ready to use My Genes.
Besides Python and the modules, you'll need to download the repository.
Uncompress the mygenes.zip file you downloaded and double click on mygenes.py under the current stable version folder.
Individuals Number: The fixed number of individuals that will figure in each generation.
Number of Generations: The number of generation that you want your simulation to run.
Node duplication probability (Probduplic): The probability of a node to be duplicated as a mutational event.
Node elimination probability (Probelim): The probability of a node to be eliminated as a mutational event. If the number of nodes in the individual is 1, then Probelim will be equal to zero until it gets more nodes.
Edge creation/elimination probability (Probalpha and Probdelta): The probability of an edge to appear (Probdelta) or disappear (Probalpha) as a mutational event.
Iteration steps: Fitness is calculated by a boolean analysis. If you're not sure how to manipulate this parameter use 5.
Replicas: The number of times you want the simulation to run. You can use the different replicas to perform statistical analysis. Recomended: 5.
Fit selection: The selection method, i.e. which individuals figure in the subsequent generations, you want to use. Currently you have the selection by fitness (0) and random selection (1).
Feel free to contact me at anytime
Daniel Vilar Jorge
d.vilar-jorge15@imperial.ac.uk
Imperial College London
Department of Surgery and Cancer