could known snp be disease causing variant?

[gary]

I am using exome seq to search for causative mutations in a family with Mendel type genetic disease.

A mutation was found in all the patients while absent among normal people of the pedigree. function of the gene is strongly supporting the phenotype of the disease.

however, the mutation has been reported in the 1000 genome project, although in a very low frequency(1 out of 1094). clinical features of the disease is very mild, only affecting skin pigment distribution.

my questions are:
Does this mutation worth further investigation?
If does, what should I do?

thanks for you attention.

[dpryan]

You can run the statistics to see if there's a significant correlation. From what you write, I would assume there is. Normally, then, you would just start with the regular molecular biology experiments appropriate to this protein and then make a mouse (if appropriate to this gene, it's not for all of them). That's pretty much the normal flow of disease research.

[gary]

thanks for your valuable reply dpryan!

The mutation was found in all the patients while absent among normal people of the pedigree. So I think it is clear that there is a significant correlation.

My primary concern is that since the mutation has been reported in the 1000 genome database (absent in the dbSNP database), is it still worth further functional investigation?

Are there any similar precedent studies I can learn from.

[Bukowski]

I think the other issue is how do you know that 1000 Genomes individual doesn't in fact have the condition? I don't think the subjects are particularly well phenotyped. I think people need to start treading very carefully with the assumption 'because it's been seen before, it isn't pathogenic'. dbSNP has plenty of pathogenic mutations..

[gary]

thanks Bukowski

Yes, that's what I'm thinking too.

Is there a way I could get the detailed phenotype info of the 1000 Genomes individual who has the variant?

[Bukowski]

I'm thinking 'no' from their not so helpful phenotype page: http://www.1000genomes.org/faq/can-i...mation-samples

[Simon Anders]

The mutation was found in all the patients while absent among normal people of the pedigree. So I think it is clear that there is a significant correlation.

You have not told us yet how many people you have in your statistics. I hope it is a reasonable sample size. (Have you calculated a p value with Fisher's test?)