Finding Papers About Well-Characterized Snps

Question

[User 6659]

Esteemed bioinformaticians

I would like to find some examples of the different types of SNP (e.g. missense, nonsense, stop gained/lost, splice site) that are associated with a particular disease and where the mechanism by which the SNP confers pathogenicity has been elucidated. I would really like the 'famous' examples if such a think exists - like the SNP equivalents of the philadelphia chromosome. If i search medline for 'non synonymous snp' or the like you can imagine how many results i get! I couldnt restrict an OMIM search to SNPs. I tried SNPedia as they only report interesing SNPs but that doesn't necessarily mean the molecular mechanism of the disease is known.

Can you offer any tips for finding well-characterized FUNCTIONAL SNPs? Or please can you share any examples you may know of. It would be much appreciated

Many thanks

Answer 1

[Casey Bergman]

One approach would be to query PubMed for your disease of interest, download all abstracts and any full text articles you can get, and then apply a mutation detection text mining system to to the resulting text. This approach has been successfully applied to finding mutations in human kinases using the Hunter lab's MutationFinder system in a recent paper by Krallinger et al.

[User 6659]

hi - i don't really have a disease of interest which is why the scope of my search is too broad for me to manage. I am just trying to find well characterized snps from any disease to simply build up a repertoire of examples

Answer 2

[Mary]

I was also gonna offer the GWAS catalog, but then I realized you wanted more concrete and specific examples.

I have another way to go: what about the GeneTests site? The reason I'm suggesting that is because if there is a specific tests (some are clinical, some are still only research tests) then they have passed a serious threshold of specific causation if they have reached the clinical test point. That said, not all will be SNP based issues--there are a variety of situations in there. So you'll still have to sift through, but it will be a smaller haystack. And they also link to HGMD.

In their "Gene Reviews" sections they also bring the relevant literature.

I did a search for "parkinson" and looked at the result for juvenile parkinson. They also link to "locus specific" in the buttons list. That goes to this page: http://grenada.lumc.nl/LOVD2/TPI/home.php at the Leiden Open Variation Database. On that page you can also link to a section that says "Summary of all sequence variants in the PARK2 database, sorted by type of variant (with graphical displays and statistics)" and has a very cool graphic page that illustrates the issues, including substitutions, deletions, insertions, frame shifts, nonsense...etc. And that appears to link to the data. But I haven't really used it before, so you'll have to explore and verify.

Hmmm...might be time for a tip-of-the-week on LOVD....

[User 6659]

thanks for the advice

Answer 3

[Larry_Parnell]

Must it be a disease? Would some condition that is not pathogenic be allowed?

A couple examples for you are the single nucleotide variant conferring sickle-cell anemia (SCA) and malaria resistance, and variations in the MCM6 gene that allow continued expression of LCT past the time of breast feeding (lactase persistance or lactose tolerant).

APOE and Alzheimer is a pretty good example (2 SNPs in a haplotype here), but the mechanism is a bit speculative.

There are some nice recent papers on PCSK9 and heart disease, which may have been identified originally by GWAS.

I would look for papers coming from Rick Lifton's group. These will be in the area of medical genomics. Jim Lupski's genome article (his own genome was sequenced in order to identify the cause of his Charcot-Marie-Tooth (CMT) disease. These will be published from 2009 onward.

Lastly, I would peruse OMIM. You are much more likely to find interesting examples when you consider monogenic or nearly monogenic diseases. SCA and CMT are examples. Or add "monogenic" as a search term in OMIM or PubMed. Add "polymorphism" or "causal variant" to the search.

[User 6659]

no - it doesnt have to be a disease. I just wanted to build up a repertoire of examples of how missense / nonsense / splice site snps translate into phenotypes. I was hoping there would be some really famous well characterized SNPs that people might be able to reel off for me. Perhaps not. This has been the most useful answer

Answer 4

[Biomed]

I would start by NHGRI GWAS catalog http://www.genome.gov/gwastudies/ and also try HGMD web site for well known diseases usch as Age Related Macular Degeneration etc. But it is important to know that there is rarely a SNP equivalent of the philadelphia chromosome. This is the nature of the SNPs being commony found in the population so it is very hard for a SNP to be the cause of a highly deleterious phenotype, such as the relationship between the philadelphia chromosome and CML.

[Casey Bergman]

Keep in mind that the SNPs in most GWAS studies included in this database are trait-associated and there is little/no evidence about which SNP is causal for the phenotype.

[Larry_Parnell]

Exactly! Casey's point is key and so often overlooked. GWAS do not uncover the functional variant.

[User 6659]

i don't really want to use GWAS as that doesn't identify the functional variant as people have already pointed out. However are there any other disease like age related macular degeneration that are known to be associated with SNPs. I could look these up. Larry's answer below has some suggestions