author: Michiel Bakker
title: Make biologists work faster with ANIMO
keywords: biological networks, modeling, ANIMO, parameter fitting
topics: Bioinformatics
committee: Stefano Schivo
started: April 2016
end: July 2016
type: Bachelorreferaat project


Modeling in biology: great potential, but could be easier

The branch of computer science known as systems biology studies the processes occurring inside living beings by means of mathematical models. Those models are then analyzed with computers, allowing expert biologists to investigate the complex dynamics of the modelled systems.

Unfortunately, mathematical models are typically out of the comfort zone of most biologists: they have studied biology, not mathematics or computer science!

We want to help biologists acquire more familiarity with the powerful tools of systems biology without requiring them to learn several books of new theories and formalisms. For this reason, we have developed ANIMO (Analysis of Networks with Interactive Modeling), a tool that hides the complexity of formal models behind a user-friendly interface, and is designed to be used directly by biologists.


ANIMO needs you!

Over the past few years, ANIMO has grown from an embryonal prototype to a complex and powerful modeling tool.

Recently, ANIMO's exposition to the biological public has been further strenghtened by a user-friendly web interface, which allows biologists to avoid all software installation processes and dive directly into the modelling work. So inexperience in either mathematics and computer science is not an excuse to avoid modelling!

However, something can still hamper the biologist who tries to build a model. When defining the speed at which biolgical interactions occur, most experts can give rough estimates without much effort. Yet it is sometimes necessary to be ensured that a more precise/different estimate of one or two parameters would not result in a biological system behaving completely differently. This type of investigation would require the biologist to spend many long hours trying different combinations of parameter settings, a job which is both annoying for the biologist and highly error-prone (which combinations did we already try? Should we have a stategy or just change something here and there randomly?). This sounds like something that could and should be automatised: here comes parameter fitting!

Many techniques for parameter fitting in both biological and non-biological context have been developed during the years, yet not all would be appropriate in the case of ANIMO models. Your task will be to select and implement the most promising technique(s) in ANIMO and make sure that they are accessible to the domain experts. This will give a powerful boost to the modelling of complex biological systems, letting the biologists concentrate with the important parts and leaving the computational aspects to the computer.

What has to be delivered:

  • Investigation of the current features for parameter setting available in ANIMO.
  • Literature overview to choose the best candidate parameter fitting methods to be added (or adapted) to ANIMO.
  • Investigation of the possibility to use the same parameter fitting methods both in the desktop and the web-based versions of ANIMO.
  • Design and implementation of the extensions to (web)ANIMO to make advanced parameter fitting features available to the biologist users.

If successful, the outcome of the project will be an extended version of ANIMO (and possibly, of webANIMO) which allows to semi-automatically fit models to experimental data, speeding up significantly the modelling process for domain experts. This will enable the everyday use of modelling in the planning, analysis and sharing of research in the biological context.

Background  material:

Additional Resources

  1. The paper