Mutual Information

products
main search tutorial faq Pfam::Home                  
JSPWiki: Dengue
JSPWiki logo
X3D-PDB
CASP7
Protein CorreLogo
Set your name in
UserPreferences

Edit this page




JSPWiki v2.2.33


Dengue


May 15, 2007

Using Mutual Information to detect co-evolving amino acids shows promise as a technique to detect protein-protein interactions. This has strong applications in RNA virus with high mutation rates. Dengue Fever is caused by a single-stranded RNA flavivirus. The Genome Institute of Singapore and Novartis Institute of Tropical Diseases has put together a great portal on the Dengue virus. Dengue Portal

Using sequence data for 160 complete genomes of Dengue published at the Dengue Portal the following network graph was created showing potential protein-protein interactions. It is important to understand that the interaction between two amino acids with high mutual information could be positive which would indicate the two amino acids or the protein surface located around those two sequence position have an affinity for each other. This would indicate that the two proteins share an interface. It is also possible that the mutual information relationship between the two amino acids prevents the two protein surfaces from interacting to reserve the protein surface for another protein.

Once the sequence positions of interest have been identified then highlighting those sequence positions on a solved PDB structure can provide additional information about the accuracy or importance of the predicted sequence position. The Major Envelope protein E has a solved structure 1THD as a trimer. The sequence positions in E that have high mutual information are color coded. It is interesting to point out that the sequence positions come from various sections in the sequence but in 3D space form a straight line along the trimer where the center E structure is flipped and rotated 180 degrees. This is an interesting indicator that the line or axis formed by these sequence position may be of important. The other interfaces are at the end of the structure and even though they are distance in sequence positions they are all approximate neighbors in 3D space.

Top scoring 50 MI pairs

Protein E structure as a trimer where the center E structure is anti-parallel


Attachments:

mi-dengue-genes-genome-50.png Info on mi-dengue-genes-genome-50.png 102108 bytes
E-1THD.png Info on E-1THD.png 137408 bytes
E-1THD.py Info on E-1THD.py 257187 bytes


Go to top   Edit this page   More info...   Attach file...
This page last changed on 16-May-2007 21:27:34 EDT by root.


home search tutorial faq Pfam::Home
For questions or comments please contact willishf@ufl.edu