Using DNA to Explore African Ancestry - Rick Kittles

Rick Kittles: Okay, thanks a lot.

I wanted to thank the organizers for the invitation.

It's always good to come back to D.C.

For me, this is where a lot of this started for me, in terms of actually applying this knowledge and working on developing these tests.

So today I'm going to talk some about using DNA to explore ancestry in African descent populations in the New World.

And when we talk about the New World, I think we forget sometimes that there were folk who were here before we got here.

Right?

So the indigenous populations of the New World, and their genetic or gene pool, there are still remnants of that in the genetics of folks today.

But after the arrival of Christophe, or whatever you want to call him, he brought with him a lot of changes that occurred in the New World.

And one of the big consequences of the arrival of Europeans in the New World was the enslavement and the forced migration, or kidnapping, actually, of enslaved West and Central Africans who came throughout the New World, right?

I mean, came to several regions of the New World from Western Central Africa.

So today I'm going to place the genetics of African descent populations in the New World into several contexts, because when we talk about genetics, I think we have to remember that this is just one tool, one piece of information, for multiple sources of information.

So I'm sure all of the speakers, the genetic speakers today, are not espousing any genetic exceptionalism.

I know I'm not.

But I do want to stress that it is a powerful tool that could inform and could allow one to reconcile individual, familial, and geographic ancestry.

So today I'm going to talk about -- place it into historical context.

We talk about the expansion and the bottleneck of the Native American component, European colonialism and the transatlantic slave trade, social/political context, because there is no monolithic African-American or African descent population.

I mean, there are different local experiences, and when we look at the genetics, a lot of that arise because of the differences in the local experiences across the New World.

And then the psychological, because folks identify themselves, and you have the right to identify yourselves how you are, whether you're Tiger Woods or Rick Kittles, you know, you say what you are, right?

A Cablasian, or whatever, right?

So when we look at these populations in the Americas, we know that there's high genetic heterogeneity, and I think that we've mentioned that earlier, due to the antiquity of the gene pool of Africans, and then also gene flow and admixture with non-African populations.

And this admixture can increase linkages equilibrium, which you'll learn more about in the next session.

And we know that the pattern of variation, of genetic variation, differs geographically across the New World.

And we also know that there's high levels of population stratification.

So when we talk about the transatlantic slave trade, we're talking about the forced kidnapping of folks from northern Senegal, southern Angola, for the most part.

Five percent of enslaved Africans in this region coming over to -- being forced over into the New World, mainly the Caribbean and South America.

Only about half a million were brought to North America.

The bulk of them, 4.5 million or more, actually, to the Caribbean and South America.

And so when we -- 10 years ago, when I really started working on this and set up the company African Ancestry, we were using the sex link markers.

Now they're kind of going out of favor, but they still were quite informative for tracing migration patterns and saying something about ancestry.

These sex link markers being mitochondrial DNA in the Y chromosome, which are paternally and maternally inherited.

This is a paper that I did with Mark Shriver where we, you know, showed the different mitochondrial DNA lineages that have emerged as populations moved out of Africa.

Very interesting stories they tell.

But they only tell one particular story out of many.

And I think we really need to understand that -- Joanna showed when you look at your ancestors as you go back in time, it can increase enormously, and so we can't tell you everything about all of those ancestors, but we can something very concretely and -- about particular lineages

when we use these markers.

So 10 years ago we started offering this DNA test for African-Americans, and we were using the Y chromosome polymorphisms and mitochondrial DNA polymorphisms. Now we have some, what we call admixture markers, which are autosomal markers, which represent a mixture of your gene pool from multiple sources in your family tree.

We call it the My DNA Mix, which I think is interesting.

You know, it's a nice name, right?

[laughter] Issues surrounding reliability of results.

And so when we talk about these lineage-based tests, you know, it's a function of how comprehensive your database is, right?

So, you know, large number of genetic data, the sampling has to be consistent with the historical records, so we have a database called the African Lineage Database, where we worked with archeologists, anthropologists, and historians to say, well, you know, what's the populations -- what present day populations are going to be important to sample for ancestry for African Americans?

And then when you talk about the accuracy of these linage matches, we have this sequence similarity measure, and then we also look at the frequency of these matched haplotypes.

I mean, the -- yeah, the matched haplotypes that we find, because, for the most part, when we look through our database, we find multiple matches across multiple ethnic groups across multiple regions.

A lot of them are single sort of matches, but then we find some that cluster, that have a high frequency in a particular region.

Well, we can come up with a statistic in terms of probability and say that's the most likely place of ancestry.

So we use those statistical measures for our matches.

Our database, currently, over 14,000 mitochondrial lineages, over 16,000 Y chromosome.

As I said, we started this 10 years ago, in 2003, so we were using a small set of molecular markers, of genetic markers, but they were quite informative for saying something about African maternal and paternal ancestry.

When we look at the genetic distances, and geographic and linguistic references in the database, we found that for mitochondrial DNA, there was a close correlation with geography.

So the closer individuals were, of course, the more they shared genetically.

And language wasn't correlated with maternal mitochondrial DNA.

So, quick story.

Jake told you earlier about the work that they were doing with looking at autosomal markers and looking in South Carolina, finding the match with the -- in the, what we call the Guinea Coast, or the Senegambia/Sierra Leone/Liberian coast area of West Africa.

We also found a similar story for the Y chromosome, and we looked at several Y chromosome samples from African-Americans in the US.

So this is that black rice hypothesis, where there were regional differences in enslaved Africans based on the principal cash crop of the plantation, and may have lead to the geographic stratification in the African-American gene pool.

So South Carolina, the principal crop in the antebellum period was rice, and we know that the Grain Coast West Africans had considerable expertise in rice cultivation, and so that many historians believe that the plantation owners in South Carolina preferred enslaved Africans from the Grain Coast.

And so this would be this area here: Senegal, Liberia, Sierra Leone.

This is what we call the Upper Guinea Coast, some call.

We had paternal lineage data from those populations, we compared them to other African populations, and we wanted to see sort of where the African -- these Y chromosomes from African-American men, and also Caribbean men, matched.

We found that about 30 to 40 percent of the men that we tested for the ancestry of their Y chromosomes had European Y chromosomes.

It's not shocking.

I actually have a European Y chromosome.

I don't tell folks at dinner all the time, but it is, it did happen.

[laughter] I'm, you know, I'm part of that 30 percent.

And that also, and I think based on some discussions earlier this week with several of you guys in here, you know, it's -- that's a part of -- that's evidence of the behavior of slaveholders, to some extent, in the history of the African experience in the Americas.

So when we look at the geographic structuring within West Africa, it's pretty high for the Y chromosome.

When we look across the different regions that we sampled from, and when we looked at the African-Americans from South Carolina, this was Columbia, South Carolina, there was close affinity with the Grain Coast populations, the Mandinka from Senegal, the Mende and Temne from Sierra Leone, and the Kru from Liberia.

So it did provide some evidence for this black rice sort of hypothesis.

And I'm glad to see that ancestry.com is affirming and -- what we found some time ago.

So when we look at these autosomal markers, these are markers that we get from multiple, you know, in our family tree.

There's no sort of clear pattern of inheritance, but they allows us to say something about your overall global ancestry.

And you'll learn more about that, how we can use these markers, to inform us about ancestry and risk for disease in the next session.

But if we look at individual's type of these markers and a principal component analysis, we can see that the red represent African-Americans in Chicago, the blue represent whites in Chicago, green are Europeans, and black are Nigerian West Africans, you see sort of that spread

of ancestry there, so across those first two principal components.

But you do see that there's significant overlap, too.

And so what they represent is that the spread of ancestry, as you've seen on the previous two talks, is quite wide.

And so you really shouldn't go home thinking that genetics is actually reifying or reinforcing race.

In fact, it's actually, if you look at it, it's actually breaking down these racial categories and showing that we all sort of have this continuum of genetic variation.

So using ancestry informative markers, we can accurately estimate individual ancestry.

And all populations, not just admixed populations, but all populations, for the most part, show variation in individual ancestry levels.

I kid Mark Shriver a lot because, you know, he has significant African ancestry; he won't tell me where he got it from, but, you know, he's significant.

You have, what, several of these so-called West African alleles, and you're, what, Irish?

Mark Shriver:American [spelled phonetically].

Rick Kittles: Oh, somebody said something today about "Black Irish."

Irish." So...

So...

[laughter] These estimates can be used to control for heterogeneity or -- in admixed populations in our genetic association studies, matching our cases and controls, and then also admixture mapping.

This is just a plot using structure.

Joanna mentioned structure as a program used for estimating ancestry.

On this X-axis are individuals, this is the proportion of ancestry, these are West Africans from Cameroon, whites from Baltimore, and blacks from D.C.

These are all self-report.

So you see sort of this homogeneity and genetic background in the West Africans.

The green represent a European Ancestry, and the -- what I call variants in genetic background for -- high variance of genetic background for the African-Americans.

On average, about 23 percent European ancestry.

So some individuals have very high levels of West African ancestry, some have very high levels of European Ancestry.

And this is very important, in particular, if you are involved in, let's say, a clinical trial or a genetic study where genes are involved in the study, you really have to control for that heterogeneity in the African-American population.

And African-Americans are not the only ones.

These are Hispanics.

Puerto Ricans, you see, the spread of ancestry is quite different -- the distribution of African ancestry in Puerto Ricans -- quite different than Mexican populations, which have a lot more Native American ancestry than the Puerto Rican population.

We look across the United States.

The average European ancestry in different African-American communities varies.

The lowest we saw was in the Gullah sea islands of the coast of South Carolina; the highest, 10 times that in Seattle, Washington.

And folks used to say to me, why do I like to go to Seattle?

It was, you know, my blood pressure goes down, folks are, you know, a bit more nicer, cabs stop for me, stuff like that.

You know, it's different than in the rural south.

But anyway, if you look at these -- the numbers, the amount of European ancestry in African-American populations is quite low compared to the urban north, and then also the West Coast.

Quickly.

Africans in the Caribbean.

We know that about 10 million enslaved Africans were brought to the Americas, about 4 million sent to the Caribbean, 1.5 million were sold in the British West Indies.

If we look at Jamaica, Barbados, and St. Thomas, we can see a wide spread of genetic ancestry; just like we see in the African-Americans, Jamaicans having a bit more Native American ancestry than the other two groups that we studied.

If we look at some islands that my post doc, who now is at University of Notre Dame, we looked at Y chromosomes, mitochondrial DNA, and also autosomal markers in these populations, and found that the highest amount of African ancestry on the Y chromosome, I mean, not

African ancestry -- those that had the largest proportion of African contribution on their paternal lineages was St. Kitts, and the lowest was in Dominica, which is really, the island has a lot of remnants of the Native American communities that were in those -- on those islands.

But for the maternal ancestry, the bulk of it being West African on those islands.

We look at the ancestry and formative markers, the autosomal markers, highest proportion of African ancestry in St. Kitts, Jamaica, and the highest proportion of Native Americans, as I mentioned earlier, Dominica, which is a very interesting island.

So if we look at South America -- one slide, please.

South America in the Cocora Valley of Colombia.

This is actually what we call the Black Belt of South America.

High levels of West African ancestry in this population, about 70 percent.

This is based on about 200 ancestry informative markers.

And about 15 percent American and European ancestry.

If we look further south, in Uruguay, vastly different proportion of West African ancestry, mainly European, and a lot of those Italian European ancestry, and about 14 percent Native American.

So this is very useful.

Even when we think about doing genetic studies in Hispanic populations, you definitely are going to have to count for that variation there in the proportion of genetic ancestry.

My last slide.

This is my last slide.

So I want to put this up -- [laughter] -- to introduce this for the topic for the next session, because what they're going to be talking about is estimating ancestry across not only the genome, but even, you know, across these chromosomes to look at local genetic ancestry.

So, in conclusion, significant heterogeneity, West African ancestry is predominant in most of the non-Spanish Caribbean.

Native American high in Dominica, and all groups have some level of non-Indigenous ancestry.

And we see evidence of sex by gene flow throughout the Americas.

One of the things we are going to look at -- [laughter] Mark Shriver: You keep talking, but walk over there.

Rick Kittles: Okay, all right, all right.

[laughter] Mark Shriver: All right, let's have the other panelists come up.

So, Rick, you can keep talking.

[applause]