Personal tools
You are here: Home Intro EvRe

EvRe

Introduction to evolutionary-research, the scientific research initiative.

 

  

In Short: Why evolutionary-research?

First of all, because evolutionary research (the field) is fascinating and important for life on this planet, as it includes topics like the origin and extinction of species, the evolution of pathogens and understanding genomic sequences. The aim of evolutionary-research (with hyphen to denote the initiative that runs this site, short EvRe) is, to support evolutionary research in general. We at EvRe use the following approaches:

  • Simulators. We pick promising questions in population genetics that could previously not be addressed due to lack of computing power. We turn them into an explicit model that is implemented by a corresponding simulator, which is run by evolution@home. Results are collected in databases and eventually published.
  • Science portal. We collect interesting links, materials and tools for researchers in the field as well as introductory materials for those who think about joining the field. This shall help you find, what you need to answer your research questions about evolution.
  • Toolbox. We collect tools (from bioinformatics over math to programming) that help to build quantitative models of evolution. Some of these we use ourselves, others have been recommended to us by scientists who use them.

So if you have a resource you think we should include, let us know. If you have a scientific question that can be answered by an existing simulator, then we would be delighted, if we can contribute to the answer by running a corresponding analysis. If you want to suggest topics for new simulators, then we are interested to hear about them, even if we do not have the capacity to build the corresponding simulator right away. But who knows, we may be building a similar model tomorrow.

 

In Detail: Why evolutionary-research?

Evolutionary theory is without doubt one of the most fascinating concepts in biology. In fact, it may be the only theory ever that is capable of providing a unifying framework for the vast diversity of biological facts that are being discovered (for a review, see here). Such a theory is important to help preserve the biodiversity of this planet, understand how new diseases evolve and make sense of the sequences found in the human genome.

Although Darwin's detailed observations helped to get the idea accepted, it was the synthesis with Mendel's theory of inheritance that got the theory going. Starting with the founding fathers of this synthesis, Fisher, Wright and Haldane in the first half of the 20th century, many have followed their footsteps in fleshing out what it means that individuals in this world inherit almost perfect copies of the genes of their parents. So evolutionary research has been going on for many decades and is being conducted in many universities around the globe. We at evolutionary-research (with hyphen to denote the initiative behind this site, short: EvRe) stand in this tradition and want to contribute to the further development of modern evolutionary theory.

Frequent speculations and undecidable verbal debates have made many scientists outside of the field believe that we already know everything interesting about evolution that the scientific method can possibly reveal. Yes, some questions can indeed not be solved without time-travel, but it is surprising how much we can learn, if we just combine a close look at what we see in our world with precise theoretical predictions about evolution. It takes a considerable amount of time and resources to develop such predictions. Thus one long-term aim of EvRe is to increase the resources available for scientific research that develops evolutionary theory.

Models of evolution easily exceed what even the brightest mathematicians can handle, if these models are realistic enough to meet the taste of typical biologists. The only way to investigate such models is to simulate them in the computer and the advent of powerful personal computers was heralded as a revolution for theoretical biology in the first part of the 90'. However, practical work quickly revealed severe limitations: Often computing demands exceeded given facilities, as many questions require too many simulations for an exhaustive rigorous analysis. Faced with this dilemma, biologists have three options:

  • Use super-computer. While few biologists can afford their own super-computer, they can compete with many other researchers for CPU-time elsewhere. This solves many problems, but is not a universal solution that works in every case, as many problems are even too big for many typical settings of this type.
  • Use clever tricks. Some biologists explore clever mathematical tricks that can reduce computing needs considerably. A surprisingly large class of problems can be addressed this way. However, large groups of other models violate the simplifying assumptions behind the math to such a degree that it is not clear how useful some of these simplification really are - unless they have been compared to full-blown realistic simulations. So in many cases clever tricks do not remove the need for complex computations altogether.
  • Limit your research. If none of the approaches above solves computing limitations, then corresponding models cannot be explored. Sometimes a way around this is to limit the study to some computationally feasible parameter combinations out of the full range of biologically interesting values (or run fewer stochastic repeats on costly parameter combinations). However the further one follows this path, the weaker becomes the scientific support for the conclusions of such work: How can one exclude that a slightly different parameter combination would not have led to radically different results? Such questions are so widespread that some biologists view simulation results in general with some suspicion. Thus, some agree with an anonymous theoretical biologist that once said: "Do not touch questions that cannot be answered in about two weeks of computing time on a PC".

This is exactly the point where evolutionary-research wants to help within its current humble limits. EvRe started evolution@home, the first global computing system for evolutionary biology to bring the power of public Internet-distributed computing to simulations of evolution in general and to individual-based models of evolution in particular.

It would be nice to say that now there is enough computing power for all interesting questions in evolutionary biology, but of course the reality is different. It is a complex processes to implement evolutionary models into simulators, distribute computing tasks and construct long-term databases that can be queried for results.

The current approach of EvRe is to implement simulators, one after another, each representing a quantitatively precise evolutionary model that helps to answer a limited class of questions from the sea of the unknown. Results are collected in databases that shall eventually be made available to other scientists - as soon as the corresponding code has been developed. Biological work that implements simulators and analyses the biological meaning of their results is currently done in parallel to technical development. The latter requires a motley mix of formal simulation methods, web technologies, databases, communication code, various programming frameworks, security technologies and above all, time. Since many key topics are interconnected and learning curves are quite steep, it is not easy to distribute development of evolution@home.  Some things could be sped up if EvRe would have a corresponding team of dedicated programmers working at the same place, but since that is currently not the case, progress at the moment seems to occur on geolgocial timescales. This does not mean that nothing is happening.

 

What else besides Evolution@Home?

The vision of evolutionary-research is wider than developing the scientific and practical side of evolution@home.

  • Communicate. The dependence on public CPU-time contributions puts EvRe into an excellent position to promote public understanding for scientific work on evolution.
  • Framework. When evolution@home started in 2001 it seemed best to implement each biological model as a different simulator with a separate executable. However, the recent experiences of Laurence Loewe with modelling in systems biology suggest that much can be gained from constructing a general modelling framework that reads in the structure of a particular model and does not require different executables for different models.  It is desirable to construct such a framework for evolution@home, since it speeds up the construction of models while at the same time reducing the number of modelling errors that need debugging. Implementing such a framework will take time. In the mean time evolution@home will continue to follow the "one model one simulator" paradigm. Once the modelling framework will be active, it will be possible to use it for a wide range of research questions.
  • Tools. EvRe collects tools that have proved to be useful for research on evolution or for programming. Your tips are appreciated.
  • Links and papers. There are many useful collections of links for evolutionary biology related papers of interest. We collect the links and papers we like and perhaps you want to alert us to something we have missed.

There are many other things we could do to promote evolutionary research and depending on our resources, we might expand the range of our activities. In Biology, there are many things that grow and with evolution many things happen slowly, so we invite you to watch this site evolve over the years. Above all we hope to raise your interest in evolutionary biology and challenge you to ...
 
...be part of the adventure!

Document Actions