Skip to main content

Digital Organism Simulation Environment

Project description

Digital Organism Simulation Environment (DOSE)

Life is fascinating and deeply intriguing. Despite so, life forms on Earth or carbon-based life forms as a group is just one form, one possible sample of possibly a whole magnitude of life. Even then, there are many aspects of life that cannot be deciphered even by examining current life forms; for example, how did chemical reactions organize themselves into biochemical pathways? How did life start? How is intelligence formed?

To answer such questions, we will have to restart our evolutionary time to the very beginning - clearly an impossibly gargantuan task. At the same time, studying biological/carbon-based life forms is expensive, time consuming and destructive. As a molecular biologist, there is no way I can examine the entire genome of even a bacteria in an inanimate state, then somehow allow it to continue living as if time had just stopped while I am examining it.

However, if I can simulate a bacteria or any life form in a computer, then I can make a digital copy of the bacterium, pull it apart to study it while the original bacterium continues “living” in my virtual world without even knowing that it had been duplicated. Many biologists thought of virtual life forms as a new way to learn about life itself. Studying of virtual life forms is known as Artificial Life and I term “virtual life forms” as “digital organisms”. There are several advantages in experimenting using digital organisms. Firstly, generation time can be much faster compared to most biological life. Secondly, it is usually cheaper to examine computer simulations than working on actual biological life. Perhaps the most important advantage of looking at life from this perspective is that by recreating life in a different medium, we are not limited to our own system of carbon-based life; hence, studying life as what-it-could-be.

Digital Organisms Simulation Environment (DOSE) is essentially a virtual world simulator for studying digital organisms. I will argue that digital organisms are considered living organisms (Koh and Ling, 2013). Despite so, being a molecular biologist by training, I have a hard time mapping components of digital organisms into biological life whenever such components are too abstract. Hence, I decided to design an artificial life / digital organism simulator that bears resemblance to biological life and ecology.

License: Unless otherwise specified, all files in dose/copads folder will be licensed under Python Software Foundation License version 2; all other files will be GNU General Public License version 3.

Documentation can be found at http://maurice.vodien.com/project-dose

Copyright 2010-2013, Maurice HT Ling (on behalf of all authors).

Project details


Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page