Genotype For Latino Population

Question

[J.F.Jiang]

Hi everyone,

Is there anyone here can tell me which kind of genotype data I can apply for Latino population? The paper said that it should be the combination of YRI+CEU

If so, the hapmap only offered the separated one rather than the combined genotypes

How can I do this kind of combination or other resources for the Latino population?

Thanks!

[Giovanni M Dall'Olio]

which paper are you referring to?

Answer 1

[Larry_Parnell]

There are several populations available in HapMap beyond the original four. Taken from the HapMap display of any genomic region:

Population descriptors:ASW: African ancestry in Southwest USA, CEU: Utah residents with Northern and Western European ancestry from the CEPH collection, CHB: Han Chinese in Beijing, China, CHD: Chinese in Metropolitan Denver, Colorado, GIH: Gujarati Indians in Houston, Texas, JPT: Japanese in Tokyo, Japan, LWK: Luhya in Webuye, Kenya, MEX: Mexican ancestry in Los Angeles, California, MKK: Maasai in Kinyawa, Kenya, TSI: Toscans in Italy, YRI: Yoruba in Ibadan, Nigeria.

perhaps one of these is better suited to your interests.

We work with a Puerto Rican population and it is distinct from Mexican Americans in haplotypes and ethnic ancestry. Thus, while you may be able to find population data that resembles your population or your interests, another population is different and is thus only an approximation, hopefully a good one.

[J.F.Jiang]

yeah thanks for your comment Actually I know that hapmap3 has added more population in the raw genotypes but there is no specific genotypes for Latino, so some papers referred to use the combination of several genotypes, so i am wondering how to combine the genotypes of two different populations?

[Larry_Parnell]

Specifically which ethnic ancestry/ies are you looking for (in order to help us point you in the right direction)? Why is the HapMap MEX population not Latino in your opinion?

Answer 2

[JC]

1000 Genomes also include some latino populations: http://www.1000genomes.org/category/frequently-asked-questions/population

[J.F.Jiang]

thanks but still not give the specific population for latino

And most interesting is that though 1000GP has advantage than hapmap, but the genotyped individuals of 1000GP are only ~50% compared with Hapmap

Am i misunderstanding that?

[JC]

Am i misunderstanding that?

Yes, you have problems understanding what is a latino population. "Latino" is a common word for people (including myself) with hispanic heritage, but in reality this cover a large and complex admixtures depending on the country, for example Cuba population was an effect of African and Spaniards, Argentina population also includes a Italians and Germans, Mexico population combines Spaniards and several different ethnic groups from mayans to aztecs, and so on. Basically you need first to define what latino population are you dealing.

[J.F.Jiang]

thanks a lot, I know that depends because of the specified country actually the one I use is the Argentina population including the Germans which should be considered as CEU right, so I am confused about this kind of mixed architecture.

[JC]

yes, Argentina population is really close to CEU in general.

[DG]

You have Mexican, Peurto Rican, Peruvian, and Colombian in that link. Depending on the population you are dealing with for your study one of those should probably work reasonably well.

Answer 3

[DG]

Chances are if you want to combine populations you'll have to use something like vcftools. Download the VCF files from 1000 genomes of your starting populations and use vcftools to merge. Then recalculate MAFs and such. I have never done this though so I don't know all of the step-by-step details.

Answer 4

[Javier2013]

If you want to use genotype data to examine your Latino population the best way to do that is to use either a Dimentionality reduction technique such as PCA or MDS or a Clustering Algorithm such as that implemented by ADMIXTURE or STRUCTURE.

I guess the paper you're referring to was using some of those techniques, because they usually work by using some population as surrogates for the ancestral population. For instance, a Caribbean population will (mostly) be composed of African and European ancestry. You can use CEU (Europeans) and YRI (Africans) to estimate the ancestral proportions for each individual in your population using the techniques I just mentioned. On the other hand, If you're working with a mainland population such as Mexico it would be better two have 3 parental populations, including some population Native American population as well.

I have some experience working with Latino population using genome-wide data.

Let me now if you will need further help.

Javier