Bioinformatics Exercises
  B.Problems. Pathway Analysis, Systems Biology
  Writer : Seyeon Weon   Updated : 10-14   Hit : 2588   Updates 

(For those who have taken the courses and want to submit for evaluation, please read the instructions linked on the table of contents page.)

  1. One of the best known pathway databases is EcoCyc ( Answer the following questions about EcoCyc:

    (a) What kind of knowledge representation method is used? Briefly explain your understanding about it.
    (b) What kind of graph layout algorithm is used? Briefly explain your understanding about it.
    (c) As a biologist, what kinds of functionalities would you want to have more in such kind of pathway databases?
    (Within 10 lines for each)
  2. Network or graph is not a static entity. It exhibits complex, nonlinear, and dynamic phenomena about which we still don’t comprehend much. This situation is the same for all other disciplines of science including biology. We may be on the verge of the beginning to understand these complex systems phenomena, even though we surely have long road to go. One of these kinds of attempt is what is called scale-free network. Briefly answer the following basic questions about network theory.

    (a) Give some real world examples for regular (lattice-like) network, random network, and scale-free network, respectively.
    (b) In the random network of Erdos-Renyi model, how does the mean length of the path between two nodes change?
    (c) The emergence of the giant component and the small-worldness may be the natural behaviors of the common networks in the real world. Can you explain the consequence of them based on some real world examples including biological ones?
    (d) A relatively high clustering coefficient seems natural for the majority of the real world networks. Can you explain it using some biological phenomenon?
    (e) Above three common properties of networks (i.e., giant component, small-worldness, and clusteredness) are necessary but not sufficient to model many real world networks. The degree distribution should also be considered. By adding the so-called Matthew effect (i.e., the rich gets richer and the poor gets poorer) to the network models, it was possible to model the degree distribution of some networks, too. Can you give some example that can be explained in this way (i.e., Matthew effect) among biological phenomena?
    (f) In scale-free networks, hubs exist following the power-law distribution. If the network inside a cell is indeed scale-free, what kinds of advantages could you expect by being scale-free? And, what kinds of disadvantages could you expect?
    (Within 10 lines for each)
  3. Kauffman’s Boolean network is an example in which we can observe the systems phenomena albeit it is a very simple model. Create a system with at least 4 different basins of attraction using the Kauffman’s Boolean network. Try to make the system as simple one as possible.
  4. Let us create an organism with only 5 genes. You can choose the interactions between the genes in anyway you like. Now, let us examine your newly created organism using Kauffman's boolean network. What kinds of attractors are observed in your organism? (Of course, you should submit the boolean rules, too.) Molecular biology so far mainly concerns with the interaction between genes. By looking at the Kauffman's boolean network model of an organism, what new insight would you get? Assume that we somehow invented an ideal DNA microarray, meaning that we can check on and off of each gene with 100% confidence. Briefly describe the design of your experiment using the ideal DNA microarray and Kauffman's boolean network as the model for what is happening inside your 5-gene organism.