Which stategy is the better one for our SNP study - (Aug/22/2012 )

Hi guys!

I am currently in a project where we are investigating the basis for radiosensitivity in cancer patients treated with radiotherapy (RT)
I have 37 very sensitive patients (severe reactions to RT, occuring in appr. 5% of treated patients) and 37 "normal" responders to RT. We have collected whole blood that we have/have not irradiated, so we can study the proteomic response to radiation. We also have extracted DNA from whole unexposed blood in which the SNPs will be studied.

The question:

We are currently planning our SNP study. We are thinking of looking in SNPs known to be involved in radiosensitivity from previous studies. The reason for studying these in our material is that we believe our material to be very "strong" in that we have relatively many very radiosensitive patients compared to other studies who "simply" collect 100 patiens "in a row" and then maybe get 10-15 very sensitive individuals, whereas we have almost 40 such individuals.

We believe our material to be too small to try to discover "new" SNPs using GWAS (which others appear to agree with) and feel it will be interesting to see how the previously studied SNPs do in our material

Now to the "twist": A professor has the "firm" idea that we first should study the proteomic response, and based on that, select SNPs in the pathways that appear interesting (I am sceptic to this)

Which approach is better?

My feeling is that there doesn't need to be any correlation between protein amount, and the function/occurence of a SNP of the protein (as a SNP can result in). Also, an upregulation of a certain protein does not mean that that is the protein in which to look for SNPs. To me, it sounds a bit far fetched. Also, with many SNPs in a gene, how to then select SNPs of interest? How to justify? With our limited material, we can't look at hundreds of SNPs, more like 30-40.

Our original approach, re-investigating previously studied SNPs justify the selection on SNPs based on just the fact that also others have found them interesting. As opposed to these studies, we believe to have a higher possibility of finding differences as we suspect our "sensitive" group to have a high frequency of "risk alleles".


Any thoughts, experiences, inputs, etc are most welcome!

Karl

The proteomics should be useful if you take known function into account - up/down regulation of stress response genes after irradiation for instance.