Two questions about BLASTing influenza...

[sez]

I will be finishing up my project by tomorrow; however, I am wondering about how common does the vaccine need to be to a strain in order to generate antibody response. Is there a specific number, such as 80%, 75%, etc.?

Also, I was wondering about a variant of this that I may be able to do if I make regionals, which is about a month after the school fair on the January 31 - measuring and comparing the right of antigenic drift between the HA and NA proteins, to see which one is faster for a particular subtype, like, say, H5N1. How would I go about this? I thought about taking the 7 prototype vaccine strains at ISD (http://www.flu.lanl.gov/vaccine/), put them through BLAST one by one and make a phylogenetic tree for each one; then taking a strain's sequences at the "top" of the tree, doing the same for the strains tracing back to the "root" of the tree. Then compare how similar each one is to the "root" to measure the rate of drift? Or is this too random to have any merit?

[deleted-71684]

This is a very interesting question in the field of immunology. To generate a specific antibody response the entire immune system of the body must become activated. This requires a specific influenza antigen (that is often a common protein that the virus uses or an amino acid motif unique to influenza) that is taken up by dendritic cells and presented as a formal antigen. When you are BLASTing the entire influenza sequence, you are not only comparing the portion of it that activates an immune response, but everything else including the virus's nucleocapsid protein, polymerases, etc. It would be difficult to say weather the similarity in your sequences actually reflects antigen similarity as well; so i would be careful with what exactly I interpret from this data.