Can You Get 'Combinatorial' Changes From Adjacent Snps

Question

[Andrea_Bio]

Hi

This example taken from a real database sequence:

Lets say you had a codon GAC and it was found to be polymorphic at the first position giving rise to 3 alleles: GAC Asp (D) AAC Asn (N) TAC Tyr (Y)

The codon is also polymorphic at the second position and the documented alleles are GAC Asp GGC Gly

Presumbably these 2 polymorphisms can occur at the same time to give rise to other combinations not listed here

e.g. AGC = Serine with change from G to A at the first base and change from A to G at the second base.

I have seen adjacent polymorphisms in sequence databases but I have never seen a database describe what would happen if both polymorphisms occured at the same time - they treat the polymorphisms as independent units. Is this an oversight on the part of the sequence databases or so unlikely to happen it's not worth catering for?

thanks in advance

Answer 1

[Dave Lunt]

I think it is very unlikely to happen in nature. To get AGC in your example above would require a recombination event falling between the first and second codon positions. Its not whether the polymorphisms occur at the same time but where they are found that is important. Each triplet you describe is a series of nucleotides physically joined together with the alternative codon found on the other chromosome. You could get recombination mixing the polymorphisms into new combinations but since they are so close together it will be a very rare event indeed.

[Jorge Amigo]

close regions tend to be in linkage equilibrium, although the meaning of "close" may vary through the genome. having adjacent polymorphisms not segregating alleles together is, as a matter of fact, very unlikely. but, you know, as Ian Malcolm states on Jurasic Park, "life finds a way".

[Andrea_Bio]

How do I accept your answer?

[Jorge Amigo]

you should see a tick box under the answer rating. click it and it will be highlighted as accepted.

Answer 2

[Jorge Amigo]

close regions tend not to be in linkage disequilibrium (I'm just entering this answer instead of commenting Dave's because I'm sure reading the wikipedia's LD entry will help you understand this issue), although the meaning of "close" may vary through the genome. having adjacent polymorphisms not segregating alleles together is, as pointed out by Dave, very unlikely. but, you know, as Ian Malcolm states on Jurasic Park, "life finds a way".

[Andrea_Bio]

Hi. I did think it would be unlikely but I wasn't sure if you could assume it was so unlikely that you could ignore it. I know that the closer together markers are the less likely a recombination event is but I didn't know if you could assume the chance of a recombination event between adjacent bases was next to nothing. But Ian Malcolm was annoyingly right on the film at least as you say.

[Jorge Amigo]

there are ways of knowing the LD values between SNPs. in fact, this is probably the most interesting feature of the HapMap project: calculating LD values and highlighting tagSNPs that can summarize the variability information of many others. you may check the LD values for that contiguous SNPs you are working with, and act consequently.

[Andrea_Bio]

Oh right, I wasn't aware that you could do things like that. Technology has moved on in the last 10 years. Can you calculate LD values for organisms other than humans at present? I presume some sort of genetic linkage data has to be available for you to calculate the LD. I'll be looking at cows

[Jorge Amigo]

LD values can be computed, of course, but you will need some population statistics for that. the idea behind LD calculation is to check whether the allele frequencies are in Hardy-Weinberg Equilibrium or not, so you need to have a certain amount of cows to infer such statistics. in case you do, the software that is commonly known for LD calculation is Haploview (http://www.broadinstitute.org/haploview), although many others (such as plink) do it too. hope it helps.

Answer 3

[Larry_Parnell]

Simply because two points of variation are close to one another on the physical length of the genome does not mean that recombination is hindered. It is certainly likely to be that way, as indicated above, but there are recombination hotspots where the recombination frequency goes quite high. The cut and rejoin of the DNA strands has to occur somewhere and could occur between the two positions you mention, but that would still be quite rare indeed.