| Project | Rearrangements in genomes with unequal content |
|---|---|
| Date | 06/12/05 |
| Version | 1.0 |
| Purpose of Meeting | Plans for semester break |
| Supervisor present | Leong Hon Wai |
Prof Leong highlighted the importance of an interim report in helping the student to organise the materials and as preparation for writing the final report for the project. He suggested that it should be completed by the end of this year.
A simple scoring function for rearrangements is to consider the score of a rearrangement on genome i to be the `\sum_{j=1}^{n}{C_{G_i,G_j} \Delta_{G_i,G_j}}` where `C_{G_i,G_j} = \frac{|\Sigma(G_i) \cap \Sigma(G_j)|}{|\Sigma(G_i) \cup \Sigma(G_j)|}`
This scoring function will give greater score to rearrangements which brings a genome closer to another genome in which it shares more common genes. However we may need to define some kind of threshold value for the score so that we only perform meaningful rearrangements.
Prof Leong pointed out that chosing the best rearrangement using this scoring function is similar to something in between the “good” rearrangements step and the itertive tree building step. He suggested that we can try to mix-and-match the three step in an algorithm and see which combination gives us better results. For example one way is to just use the “best” rearrangements algorithm, anther possibility is to first use “good” rearrangements and when there are no more “good” rearrangements, change to “best” rearrangements. Similarly we can also study an algorithm which uses only the tree building step as I suggested previously. By looking at different combinations we will also have a better understanding of the suitabilitiy of the different approaches.