Was this page helpful?

Cellular Automata (for Biology)

    1st of July 2009 - David La.  Seems like I will be discussing Cellular Automata.

    What's Cellular Automata?

    From the wikipedia: A cellular automaton (plural: cellular automata, abbrev. CA) is a discrete model studied in computability theory, mathematics, theoretical biology and microstructure modeling. It consists of a regular grid of cells, each in one of a finite number of states, such as "On" and "Off". The grid can be in any finite number of dimensions. For each cell, a set of cells called its neighborhood (usually including the cell itself) is defined relative to the specified cell.

    Make sure to also check out a really cool example: Conway's Game of Life

    Some good books (I guess they are available at the library or you can just contact me for a copy):

    1. Wolfram. A New Kind of Science‎.  (2002) pp. 1197
    2. B. Kier et al. Cellular automata modeling of chemical systems: a textbook and laboratory manual‎.  (2005) pp. 175
    3. Deutsch et al. Cellular automaton modeling of biological pattern formation ...‎.  (2005) pp. 331
    4. Keedwell and Narayanan. Intelligent bioinformatics: the application of artificial intelligence ...‎.  (2005) pp. 280 ) (Chapter 10 is good.)
    Was this page helpful?
    Tag page (Edit tags)
    • No tags

    Files 3

    FileSizeDateAttached by 
    It's up to you if you want to read it. I recommend it.
    1286.38 kB18:09, 22 Jun 2009davidlaActions
     J Theor Biol 1993 Ermentrout (2).pdf
    You might want to read this.
    1359.57 kB18:07, 22 Jun 2009davidlaActions
    It's up to you if you want to read it. I recommend it.
    257.3 kB18:11, 22 Jun 2009davidlaActions
    Viewing 6 of 6 comments: view all
    questions are very welcome. edited 15:29, 26 Jun 2009
    Posted 22:48, 25 Jun 2009
    1. Review paper, 4.1 section fibroblast aggregation, does the shape of the "dish" affect the fibroblast aggregation? As for the size of the "dish", does it determine the "large enough" cell number?

    2. Since CA is a subcategory of AL (Artificial Life), can you give a very very very brief intro about AL, :)

    3. What's the normal way they validate their model? Use ants trial as an example, how do they know it's a correct model?

    4.What is the correspondence between CA and continuous systems?
    Continuum descriptions may be given of many of the large-scale structures that occur in CA. Presumably in the limit of large spatial dimensionality, this approximation should become accurate. But in one or two dimensions, it is usually quite inadequate, and gives largely misleading results. Are there High-Dimensional CAs? I guess those are also computationally expensive edited 22:21, 30 Jun 2009
    Posted 21:09, 29 Jun 2009
    1. It seems that if a model is too simplistic then it does not have biological relevance, whereas if the model is too complex, then a detailed exploration of all the parameters is unfeasible due to time and memory constraint. How do we know that we have struck right balance with parameters and still ensure meaningful biological results?

    2. Also, there might be many cellular automata modeling a biological process using different parameters. What process do we use, if we want to compare the different models? Don’t we need to have a uniform system-dependent methodology for comparing models?

    3. Can cellular automata be used to study emergent properties of a system?
    Posted 18:02, 30 Jun 2009
    reazur, answer your question 3: Yes, just refer to the 2005 paper
    Posted 19:32, 30 Jun 2009
    1. Authors claim that CA can be used to get the qualitative analysis (preliminary) before doing the real world mathematical modeling. But I think initial CA analysis should be done with caution because during discretization of the state space parameters ( which are normally bounded by certain rules/values) and using 2-D CM models may sometimes not yield a complete qualitative picture (hypothesis). Is it guaranteed that CM models always agree qualitatively with mathematical models? Are there instances of differences?

    2. In deterministic CA, model derivation can be done in two ways. 1. simply abstract the observed phenomenon into few simple rules, without any analogous to full mathematical model 2. directly discretize the mathematical model.Which of these approaches is better. Is it more problem specific or is one approach performs better than the other always?
    Posted 09:11, 1 Jul 2009
    The 2005 paper only shows cellular automata applied to one signaling pathway where they model the behavior of 3 substrates and 4 enzymes.

    In a real biological system, if we wanted to see the effect of emergent properties of related pathways, I think it is becomes a more difficult problem because now you have to model different pathways at the same time.
    Posted 14:53, 1 Jul 2009
    Viewing 6 of 6 comments: view all
    You must login to post a comment.