What Is The Difference Between A Snp And An Infrequent Variation?

Question

[Andrea_Bio]

Hi

I'm returning to this field after a long haitus so please forgive the basic questions. Everything is confusing me right now as I can't remember anything and what I do remember I can't remember if it's right or not!!

What is the difference biologically between a SNP (found in 1% or more of population) and an infrequent variation (found in only a few individuals and isn't sequencing error)? Can you assign any significance to a variation that has a higher prevalence in the population? Can you say that because the SNP is found in more individuals it is more likely to have a phenotypical effect? Or perhaps you can say the infrequent variant is more recent that the polymorphism as it isn't as prevalent but may well become more esatablished in the population after more generations. Or perhaps you can't say anything at all.

If you find that a particular sequence variation occurs in many individuals in one population but is infrequent in a different population, can you make any conclusions about that?

If I wanted to understand more about this, what sort of area should I be reading about? Is it population genetics or evolutionary genetics or both or something else entirely?

Many thanks

[Michael]

Cool, this is an example of how very good questions lead to excellent answers. 1 up

Answer 1

[Larry_Parnell]

I'll address these point by point.

What is the difference biologically between a SNP (found in 1% or more of population) and an infrequent variation (found in only a few individuals and isn't sequencing error)?

Essentially none - a variant at 0.05 minor allele frequency (MAF) in a population may be called private.

Can you assign any significance to a variation that has a higher prevalence in the population?

Yes. But what significance? This depends on the analysis and questions one puts to the information. A large MAF may be indicative of selection or genetic drift. It may be meaningless if other populations have similar or different MAFs for that marker. This is a complicated topic, actually.

Can you say that because the SNP is found in more individuals it is more likely to have a phenotypic effect?

No. There is something called the "hitchhiker hypothesis." The SNP could be coming along for the ride and not exert much or any phenotypic effect. Or, we don't know how to measure that effect yet...

Or perhaps you can say the infrequent variant is more recent that the polymorphism as it isn't as prevalent but may well become more established in the population after more generations. Or perhaps you can't say anything at all.

Exactly, we may not really know. A SNP with low MAF may have had higher MAF in the past and that minor allele is now being selected against. Or that less common allele has just emerged. There are tests that one can do to try to distinguish this. For example, is the SNP found in other populations, and which ones. Still, it may be difficult to assign significance to a SNP with MAF of 0.012 in one population and 0.003 in another.

If you find that a particular sequence variation occurs in many individuals in one population but is infrequent in a different population, can you make any conclusions about that?

Basically, it could be due to selection or it could be a founder effect. Read papers about the HapMap populations or from those who are looking at population admixture - see papers in this regard by Lai, CQ, or Bustamante, C (regarding NextGen sequencing or genome-wide patterns of population structure and admixture) or Burchard, EG (asthma in Mexican children, ancestry as a modifier of pharmacogenetic gene-gene interactions, and asthma in African Americans).

[Istvan Albert]

What a great answer, it covers so many issues. Feels like listening in on a lecture on SNPs.

[Larry_Parnell]

Thank you, Istvan! The questions were posed in a logical manner, which helps to give the impression of the "lecture." Nonetheless, I've been using SNPs in my research for a decade and so I know this well.

[Andrea_Bio]

Thanks for a great answer Larry. I have tonnes more questions now if you are interested. I don't know if participants get any bonus or rewards (other than altruism!) from answering questions. I could ask the related questions here but you don't get much room in these comment boxes.

[Khader Shameer]

Larry: Thanks a lot for sharing your insights on the topic. It will be great if you may please add the link to the specific papers you have in mind.

[Andrea_Bio]

I second the request for the specific papers :)

[Larry_Parnell]

To andrea_bio: If you have other questions, send them by email to larry dot parnell at tufts dot edu.