Snp With More Than One Base ?

Question

[Pascal]

If I remember well some genomics lectures I had, there were some SNPs with more than 1 base?! Can you confirm it is possible? In which situation does it happen? Could you give me some reference ?

[Michael Kuhn]

What does the abbreviation "SNP" stand for?

[Pascal]

It stands for Single-nucleotide polymorphism. More information here: http://en.wikipedia.org/wiki/Single-nucleotide_polymorphism

Answer 1

[Dave Lunt]

Hi Pascal if I interpret your question correctly you are saying that you learned that there are some Single Nucleotide Polymorphisms (SNPs) in the human genome that are tri-allelic (3 different nucleotides in the population) rather than the much more usual bi-allelic. It is better to use this terminology (bi-allelic, tri-allelic) as a SNP cannot have a 'single base'.

This is true, though the reasons are not well understood as far as I know. You could start with this paper if you want more reading

Alan Hodgkinson and Adam Eyre-Walker

Human Triallelic Sites: Evidence for a New Mutational Mechanism?

Genetics

January 2010 184:233-241; published ahead of print November 2, 2009, doi:10.1534/genetics.109.110510

http://www.genetics.org/content/184/1/233.short

Answer 2

[Pierre Lindenbaum]

if by one base, you mean "the variation occurs on two or more base": yes, dbSNP contains some SNP having more than one base:

+-------------+-------+------------+----------+---------------------+
| name        | chrom | chromStart | chromEnd | observed            |
+-------------+-------+------------+----------+---------------------+
| rs57613521  | chr1  |      14062 |    14065 | -/ATC               | 
| rs4030206   | chr1  |      14719 |    14722 | -/GAG               | 
| rs3982657   | chr1  |      15590 |    15594 | -/TTTG              | 
| rs71273311  | chr1  |      17364 |    17373 | CGTAGGTAC/TGTAGGTAG | 
| rs71274494  | chr1  |      17451 |    17479 | lengthTooLong       | 
| rs71283235  | chr1  |      20220 |    20227 | ACTGTGC/GCTGTGG     | 
| rs3039165   | chr1  |      20564 |    20567 | -/TTG               | 
| rs71309401  | chr1  |      21346 |    21352 | ACGACA/GTGAC        | 
| rs71301114  | chr1  |      21872 |    21874 | CA/TG               | 
| rs113122596 | chr1  |      24347 |    24349 | -/CC                | 
+-------------+-------+------------+----------+---------------------+

dbSNP was the "single nuclotide polymorphism database" but it has recently been renamed: "Short Genetic Variation database" : http://www.ncbi.nlm.nih.gov/projects/SNP/

[Pascal]

Thanks Pierre. So in this case (= in the screenshot of dbSNP you inserted in your post), where there are more than one base observed, can we still call it a SNP in the sense of "Single Nucleotide Polymorphism"?

[Larry_Parnell]

You could call it a SNP, but better to use terminology that includes "variant" or "variation." In conversation, SNP would suffice, but would be incorrect in a presentation/publication.

[Mary]

FYI: UCSC browser gets around it by calling it "simple" instead of single.

Answer 3

[Aaronquinlan]

I interpret your question to be about multiple, consecutive nucleotide differences, not multi-allelic variants that affect a single base pair. If so, these are often called MNPs (Multiple Nucleotide Polymorphisms). While rare, they do in fact exist. The 1000 Genomes Project is working to specifically annotate these. As Pierre illustrates, there are examples already in dbSNP.