Causative Snp Data Set

Question

[Abdel]

Hi All which is the best database choice from where i can extract a data set of causative variants and a data set of benign variants (OMIM ,GWAS). Thanks

Answer 1

[Sean Davis]

You might take a look at the references for SIFT, PolyPhen, and other tools like them for the datasets they used.

Answer 2

[ff.cc.cc]

Hi, you might also take a look here:

Harvesting Known Disease Mutations

Answer 3

[Larry_Parnell]

+1 to Sean for a good answer - this relates well to coding SNPs.

One of the first causative variants with regard to human disease to be described was the Philadelphia chromosome leading to certain forms of leukemia. Later analysis showed this to be a fusion between BCR and ABL genes. The hemoglobin variant leading to sickle cell anemia is known. This also leads to some resistance to malaria. OK so far, these are all diseases, or are we not OK? It is not clear from the question if you're only interested in SNPs or any kid of variant. What about trisomy 18, for example?

My point here is one of shades of gray. I've presented above some clear cases of disease but the variants are all quite different from one another. To show some shades of gray on the disease side - what about blue eyes and fair skin and OCA2 variants? Do these increase risk for skin cancer and vitamin D levels (some say yes), and do they cause those conditions (harder to say so definitively but there may be interesting links via immune pathways). In other words, it is difficult to know if a variant causes disease or just increases risk for that disease where the disease manifests after a chain of events.

A long preface to say - be careful on how and which data you collect. That said, I'd use OMIM for tighter relationships, clinically relevant SNPs from dbSNP also for good disease relationships, and GWAS for suggested disease relevance. As an aside, there is a GWAS hit at OCA2 for Parkinson's disease (age of onset). I am not against GWAS but believe that these tests only show association and not causality. The causal SNP may be in high linkage disequilbrium (LD, where r[?]2[?] > 0.9, for example) to the GWAS SNP or the GWAS SNP is hitchhiking with the rare, causal SNP.