Report of the American Society of Human Genetics Ancestry Inference Roundtable (Panel)


Malia Fullerton:
With this — do you think it will work? So — is this on? Does this sound like it’s on?
Yes, it is on. Okay, great. So, thank you so much Charmaine for kind of
doing the quick overview. The moderator of the last discussion opened up questions to
the panelists of each other. I’m not going to ask you guys to do that. That wasn’t the
plan. But, rather, we thought we might actually start by trying to bring all of you in the
room a little taste of the kinds of conversations that we were having as part of this roundtable
discussion. Now, I know that a certain proportion of you in the audience were actually participants
in the roundtable, but not all of you were. And so I want to ask you, for a moment, to
put yourselves in the shoes of someone who might have had a chance to participate in
this kind of multi-disciplinary, multi-stakeholder-type conversation that Dr. Royal described, where
you’re coming with your particular interests, your particular points of view, and you’re
being tasked with coming to consensus about the meaning, and uses, and applications, and
potential regulation, and things like that, of this interesting, exciting, but possibly,
as was noted by the last set of panelists, not all that well-understood yet, science. So one of the things that we ended up spending
a lot of time talking about was just, what is this thing, ancestry testing, right? So
probably all of us have some sense of genes and genetics, and we saw images of double
helices, and we saw pictures of families and resemblance, but what is ancestry in the context
of genetics? And what is ancestry testing? When I posed that question — and again, I’m
not going to ask these guys because they were there, and I was there — but what do all
of you — what would be your answer? If you were someone who was there and you were trying
to reach consensus with people, what would your answer to that question of what ancestry
testing is, what it means to you? Now that you’ve been here all day, how do you think
about this technology in the light of the sociocultural, and historical, and artistic,
and other types of implications and meanings that we’ve been talking about? Please, anyone. [laughter] Microphone. No? But you all had — every time
we asked someone to ask a question earlier, everyone — there were people jumping up. [laughter] Yes, please. Female Speaker:
All right. We’ve got a couple people. Yeah, great. Sandra Urahli [spelled phonetically]:
What I have — I’m Sandra Urahli from Trinity Washington University. I have family from
Mississippi and I have become interested in knowing more about the family history. So
I have actually — it was interesting to see some of the photographs of the 1870 census
because I was able to find that, and I found my great-grandfather in those records. And I had been interested, and have been interested,
in trying to figure out how I might have some genetic testing done, sort of to give us a
better idea of our genetic origins. But as I listen to the conversations today, and some
of the concerns about the reliability of those data as presented, I’m not sure that it’s
going to provide for me the kind of information that I want, at the level of accuracy that
I would like. Malia Fullerton:
Okay. So, I hear you saying “intriguing,” but maybe won’t be giving you the answers
that you’re looking for. Is that paraphrase about right? Yeah? Yes, what would you say? Female Speaker:
Hi. My question would be — well, my answer would be, it depends on the context of how
you use it. And if a person is using ancestry in order to have a deeper meaning to understand
themselves, or if you’re using it to connect you to someone else or another group of people,
or you’re using it from the position of trying to classify other people — it depends on
the context of how you use it and what information you’re looking for from it. So if you’re looking to be more precise, as
I guess we were able to see here, that the precision isn’t what some would like, and
for others, it’s a lot more than what we’ve had before. So I will stand by, “It depends
on the context.” Malia Fullerton:
Right, so, context-dependent. Anyone else what to venture an opinion? Mike, you haven’t had a chance to speak yet.
Let’s pretend like you weren’t there. [laughter] Michael Bamshad:
So, let’s see, is this on? You know, I think, from our perspective, ancestry testing, and
even more specifically, genetic ancestry testing, can be defined rather broadly. So one of the
first questions, and Charmaine mentioned that what we did is split up into smaller groups,
and then each of the small groups came together for a larger group discussion. In many of
the small groups, in fact, one of the first questions that was asked was, does a genetic
inference of ancestry depend upon doing a test with DNA? So there are many types of inferences, even
genetic inferences, of ancestry that one can make from sources other than DNA. In fact,
you can get a blood test and look at a cholesterol level, and that could give you some information
about ancestry. Clinicians, particularly clinicians who focus on medical genetics problems, might
be able to look at an individual and make an inference about their genetic ancestry. So that was actually one of the first topics
of discussion in our small group, was whether you actually had to do a test with DNA, in
other words, get a direct estimate of genetic ancestry, in order to constitute a genetic
inference of ancestry. So that’s one of the points that we struggled on. Malia Fullerton:
And then there were also questions, weren’t there, Charmaine, about which sort of ancestors
one might be able to access via genetic test, right? Charmaine Royal:
Right. Malia Fullerton:
So we have people who might be interested in understanding their heritage going to the
18th century, for example. There’s some people who might be interested in ancestry of thousands
of years ago. And there were other people who were interested in understanding about
near relatives, people whose family relationships have been disrupted in some way and are looking
for connections. I don’t know if you have any other comments.
Before we do the reveal — Charmaine Royal:
No. I mean, people — no, go ahead. [laughter] I was going to say, people do go into this
for various reasons in terms of what they’re looking for. As someone said, it may not provide
her with what she’s looking for, but for other people it’s very informative, and they really
rely on the information. And we’ll talk a bit more about that kind of reliance on genetics. Female Speaker:
One of the things — I’m a genetic genealogist, and I was at the conference, and I’m the president
of a large genealogy convention — association, with 2,200 members. There’s a lot of confusion.
“Ancestry” is the broad term. “Genealogy” is looking for specific individuals. And oftentimes
we’re using “ancestry” when we’re meaning “genealogy.” And so, there’s a relationship,
certainly, but there are some differences. Malia Fullerton:
Yes, and you’re capturing some of what we spent time talking about, was all of these
different ways in which ancestry could be understood. Michael Bamshad:
Right, and if I could add to that. You know, one of the other things that we still struggle
with, despite the consensus statements that we’ve come up with, is that you can — you
can define ancestry in many different points in time. So it’s very easy for us to think
about ancestry when we’re thinking about parents and our grandparents. And it’s, perhaps, easy
to think about our ancestry when we talk about modern human origins emanating from somewhere
in East Africa. But, remember, anything in between those two points is also ancestry.
And we arbitrarily decide where to define what we’re looking at, back in terms of time
and space. And that’s where we have much more difficulty when we talk about ancestry, because
often, we’re talking about different points in time and in space. Linda Heywood:
Yeah. I would like to deal with the question of admixture because now that’s a big part
of this sort of cutting-edge, that you can get more refined, you know, information with
the admixture tests. And for my particular admixture test, the first set of results I
had indicated that I had, in the 70s, African — 17, or so, to 19 percent Euro, American,
you know, white. And then, something like 3, or so, percent Native American. Now, I had grown up with the story in the
Caribbean that my grandmother actually was Carib. So when my kids were growing up, I
actually, you know, told them the story about their Carib ancestor, and Native American,
that would be in America. And as I got the test, and my younger daughter, who’s now a
doctor, she has a fellowship in infectious diseases, grew up with that. So when she was
at Harvard as an undergraduate she actually — she did a course on Caribbean history,
and — with Orlando Patterson — and, in fact, decided to do her paper on the Caribbean,
you know, on Native Americans in the Caribbean. Of course, she had grown up with this story
from her mother about, you know, this Native American ancestry. I had to sort of divest of that story when
I found out that it was more the European mixture and the Fulani mixture in my, you
know, ancestry that may have been responsible for the non-African. And just the — every
time — it would seem to me that the Native American sort of presence seemed to be shifted
a lot. And I don’t know whether it’s because of, you know, enough data hasn’t been — but
that’s where the big problem is, I think, in the Native American, because a lot of,
even in America, a lot of Afro-Americans have stories of their Native American ancestry,
and then when we get the results, that’s not it. Can you sort of, you know, explain some of
the problems with admixture testing? [laughter] Michael Bamshad:
Problems. So, let me try to put that in the context of what, of the comment that I just
made in the sense that, to a large extent, and particularly in the United States, we’re
all admixed. But whether you’re admixed or not, and to the extent to which you’re admixed,
and the populations that served as the source for that admixture, are highly dependent on
assumptions you make about the type of DNA genetic — the type of DNA test that you’re
doing, your reference populations. And that goes to where you’re making that line, back
in time and in space, and deciding, essentially, where to look. So there are a lot of assumptions that are
made in virtually all the tests that influence whether or not you get a particular percentage
of admixture from one population or another, or are considered admixed at all. And you’ve
heard some of those. I mean, one of the consistent themes through this day has been that your
ancestry estimations are highly dependent upon the reference populations that are used.
And right now, we have only a very limited sampling of worldwide populations. So if the
population from which you have admixture simply isn’t represented in that sampling, then you’re
going to get an estimate from the next closest population. And that next closest population
could be quite different than the real population from which your admixture rose. But there
are a whole host of assumptions and different data sets that are used that go into these
admixture estimates. And, just for a moment, that kind of gets
to something that was brought up by the last set of speakers, and that they were surprised
that science is gray. [laughter] You know, no bad pun intended, but science
is virtually never black and white. It’s always gray. And not only do we not always have the
answers, if we had all the answers, we wouldn’t be doing science any longer. Good science
means that you try to answer questions, and in the process of doing so, you raise 10 other
questions that you move on to answer. And, in fact, often you prove, initially, the question
that you asked is the wrong question to ask, so… Charmaine Royal:
Yeah. I think that the conception, or, yeah, or the idea that science is black and white,
or the idea that science is always right, in part has to do with how science is presented,
how scientists present their science, how journals present science. And so I think there
has to be — a lot is done with the interpretation of the information that leads people to believe
that science is supposed to be a particular way, or a finding means a particular thing.
I think a lot of it has to do with how it’s translated from the findings to the public
or to other scientists. Malia Fullerton:
Yeah. And, you know, and I think also, in this domain in particular, not only do we
sometimes think that science is black and white, but we often — which is why it’s so
wonderful that we just had the panel before. We often think of science as being locked
away in the lab and separate from us out here in the rest of the world. But particularly,
I think, in this area of ancestry testing, I mean, it was just abundantly clear in our
discussions, the people who are engaging in this testing, who are interacting with one
another in social media, who are coming back to the laboratories and saying, “But you gave
me this result last week, and now you say the result is this,” that they are actively
co-producing information about what ancestry is and what ancestry means together with scientists.
That this isn’t about the scientists in the lab saying what ancestry is, but it’s all
of us collectively together. And I was watching that going on in this roundtable. It was really
remarkable. Michael Bamshad:
Yeah. I know you had a question, so I just wanted to add that science is very much a
social process. It’s not done in isolation. The questions that we ask are often questions
that we think are going to be important, are going to be fundable. And right now, I’d go
so far as to say that the African-American community is certainly driving a lot of the
most interesting questions that are being asked about, not only ancestry testing, but
population genetics and human genetics in general. Malia Fullerton:
Yeah. No, that’s a great point. Karen [spelled phonetically]. Female Speaker:
So, this isn’t really a question, but maybe an insight from the day. First of all, you
said — you asked the question about ancestry, so maybe I’m a product of Alex Haley’s series
when I was growing up or — but, you know, to me it was roots. And, you know, of course,
that’s contextual, how you look at your roots. But I wanted to pick up on — just pick up
on where you left off. The — probably the greatest insight from the day is that ancestry
testing, I think, we see as objectifying, or sort of validating a story. So, you know,
you have this story from your family, and then the ancestry testing is either going
to validate it or raise new questions. And why have we privileged one over the other?
Because maybe the story is right and the test is wrong. Or maybe it’s a little bit of both. And what the insight from the last panel with
art, was I thought, boy, the art’s there to objectify some of this stuff that didn’t have
a context. You know, I mean, looking at the picture. Like, “a picture’s worth a thousand
words” kind of thing. So I thought the day was — there was a level of humility among
the scientists, and a level of humility among the rest, perhaps as well, to realize you
can’t have one without the other. I mean, in order to give this context, you need to
integrate it with what you know about the stories, what you know about what they look
like, what you know about the interactions, what you know about the social context. And
— but you can’t look at any of them in isolation. So, I don’t know. I don’t know. I wasn’t at
your reveal you were making. Malia Fullerton:
Well, but I mean, I think that was a big take-home. Yes, thank you. Michael Bamshad:
But I’d say it even more explicitly in a sense that, you know, you saw that there was a plot
that showed, kind of, what individual samples from across Europe, where they — how different
they were from one another. And the plot looked very similar to where they were sampled from
Europe. Remember, that if we didn’t know where those individuals came from, so if we didn’t
know something about their history, we wouldn’t know how to interpret that plot. Malia Fullerton:
They’d just be dots. Yes, yeah. Michael Bamshad:
So, in fact, without history, without anthropology, without the social sciences, we couldn’t interpret
these genetic data the way that we do. And so they’re very much a process that are intertwined
with the other sciences. Female Speaker:
[inaudible] Malia Fullerton:
Yeah, very true. Yes, please. Male Speaker:
I have a question about one of the bullets that you presented early on that said something
to the effect that population genetics need improved, quote, “enforced standards,” end
quote. And I — part of my background has to do with — is in the forensic sciences
community. As you may know, that community has faced, and is facing, this question about
— and I mean the scientific societies — about how to establish and “enforce,” in quotation
marks, standards among their members. I would really like to hear from you folks about what
sort of discussions touched on that aspect of this problem. Michael Bamshad:
Well, we did have a small group and large group session devoted to this issue. And we
actually were able to achieve some consensus, that consensus being that it would be helpful
to have some standards against which ancestry testing could be assessed. That is, making
certain that the data, the methods, the models that are used are described to people who
are using ancestry testing. And this is true for researchers doing ancestry testing as
well as direct-to-consumer ancestry testing companies. And that this information is transparent,
at least as transparent as can be. Ideally, some of this — all of this information will
be publicly available. But, for example, some of the — some of the information a researcher
or company might use, in terms of what are the — technically, what are some of the — what
is some of the information about the populations that they use as reference populations, that
could be considered proprietary. So we actually did agree that some standards
would be helpful, not only to consumers and participants and researchers, in research,
but also to the individuals who are doing the work, in other words, the researchers
and the companies themselves. Because one of the concerns that we had is that the credibility
of the research or the credibility of the product coming out of a company could be undermined
if people weren’t confident that the results that they were getting met a certain set of
standards. Male Speaker:
I guess I was more — I was more interested in the enforcement part, because that’s something
— I don’t know if your society has already had experience with various means of enforcement,
such as accreditation, or whether that was something that you discussed during your meeting. Malia Fullerton:
Charmaine, what? Charmaine Royal:
No, I was just going to say, we didn’t get into the nitty-gritty of what enforcement
meant or what it would look like. So our plan — we developed these consensus statements,
we completed them yesterday. We’re going to continue working on them and refine them,
and likely, we’ll publish them at some point. And those statements will lead to the guidelines
that we talk about developing. And when we get to that point, we’ll really
come to decisions about what enforcement will look like. And those decisions will be made
in concert with researchers and companies, because it doesn’t make sense for us to lay
down the law about what enforcement should be if it’s not going to happen. So we meet
those stakeholders at the table, too, in terms of deciding how we will enforce these things.
So we’re going to — we hope to get to that point. Malia Fullerton:
Yes, thanks. Michael Bamshad:
So I was going to say, you know, one of the things that I heard discussed was actually
some of the users of the ancestry testing companies are also — they’re also very interested
in standards, because they want to know that the information that they’re receiving is
reliable, and doing something like a seal of approval. So, such a seal might motivate
a company to — you know, they might realize that it’s to their benefit to have such a
seal because then their product might look more attractive to a potential buyer. Malia Fullerton:
Yes, please. Female Speaker:
Hi. I wanted to know if you could elaborate on the bullet point about federal regulation
possibly not being needed. When I read that, what immediately came to mind was a bill that
recently died in a legislation in California, where the bill was basically proposing that
donor consent was needed for the genetic testing and analysis of, I guess, you know, body — any,
you know, hair parts, or, you know, skin that is obtained, you know, by chance. And so in
— along those lines, it made me think of, like, genetic surveillance. And so, if we
don’t have, you know, any sort of regulation on a federal level, what are the possibilities
of, you know, in a day and age of, like, hacker spaces and bio-hacker spaces, where people
can, you know, have easy access to kits that will allow them to extract DNA from random
hair fibers that they’ve come across and use any sort of online database to make any sort
of, you know, referencing. Malia Fullerton:
Yeah, well, would either of you like to — we didn’t talk about that explicitly as part
of the roundtable, but, I mean, thoughts? Charmaine Royal:
And there are some people who have recommended or suggested that some entity, whether it
be the FDA or the Federal Trade Commission, needs to regulate ancestry testing. Well,
there’s been a lot of discussion ongoing about the regulation of direct-to-consumer genetic
testing more broadly, and really focusing a lot on health-related genetic testing. And
there are questions about what’s going to happen in that arena. And so the likelihood that the FDA, or the
FTC, or some other organization will take ancestry testing on and regulate it is pretty
slim, I think, at this point. And one of the things that we thought is, well, let’s see
if there are alternatives to federal regulations, considering that it may not happen, but alternatives
in terms of getting — of encouraging the kind of behavior that we would like to see.
And then if that doesn’t work, then maybe we need to think about federal regulation.
But I think we — in the white paper we thought of it as a step-wise process, in terms of
let’s start here and see what we what we can come up with in terms of guidelines that the
stakeholders agree on, will be feasible and practical, and that they will adhere to before
thinking about going further. Michael Bamshad:
And just to be clear, that, you know, we were specifically talking about federal regulation
of genetic ancestry testing. So there is federal regulations for clinical genetic testing,
there is federal regulation about consent for use of samples, there’s federal regulation
about storage of healthcare-related data. So there is — for many of the issues that
you brought up, there is federal regulation that provides oversight. We were narrowly
focused on ancestry testing. Malia Fullerton:
Yeah. So we’re — I see that we’re getting close to running out of time. What we wanted
to do was kind of open this up and see what you all thought, and then to kind of just
give you an example of one of the 24, 25 consensus statements that we came up with. And so — and
we probably — we had a few others back pocket, but we’re probably just going to give you
this one. So when we had discussion, and recall, we
did this by breaking into small groups and having discussion in our small groups, and
then coming back together as a whole to have further discussion, we agreed, ultimately,
as a group, and then voted anonymously, to ratify this particular statement of our understanding
in the roundtable of what genetic ancestry inference means. And we said, “Genetic ancestry inference uses
regions of the genome that inform us about individual, genealogical, population, and/or
geographic ancestry by comparison of similarities to a reference sample or samples. There are
many different types of ancestry tests that use Y-DNA, mitochondrial DNA, autosomal, or
the X-chromosome, or X-chromosome markers, information about parentage, kinship, and
identification that can be captured by ancestry tests and vice versa, and, importantly” — and
this is something we spent a fair amount of time talking about — “this information can
also be used for forensic applications.” And I know that this has come up obliquely
a couple ways today, but I think it’s important for us to realize, as we kept kind of pointing
out, the very same technologies, the very same information, can be used in these very
diverse ways, all of which can tell about ancestry and can tell us about other things,
including possibly identify samples at a scene of a crime, for example. So, there you go. [laughter] [applause] I don’t know if people have any final thoughts
or, Vence, as you’re coming up, because you were a participant in the roundtable, if you
have any reflections or thoughts on the roundtable and its relation to the discussion we’ve been
having here today. Vence Bonham:
I’m looking forward to seeing the publication, and to hear the conversation with the industry
and with the different constituencies. And just like today is just the start of a conversation
and exploration, and I think from the perspective of the National Human Genome Research Institute,
we are particularly interested in it. And both from the perspective of how it helps
to communicate information and enhance the literacy of the public, but also the broader
issues of how we understand genetics and genomics, and what it means about our identity and who
we are. So I think we’re all just beginning today. Female Speaker:
Yes. Vence Bonham:
But I know it’s 6:00, and we’re actually going to stop on time. And I want to just thank
everyone for today. And, again, on behalf of the planning committee, I want to thank
you for your commitment to be here all day, for those that are here, and for the conversations
we’re having. There is the event this evening that starts at 7:30, where Ms. Gwen Ifill
and Mr. Lonnie Bunch will have, revealed by Skip Gates, some information about their history,
their genealogical history, as well as information shared about their genomic history. And then
we’ll continue the conversation about what that means, and what that doesn’t mean, and
how do we think about it. I invite you all. If you don’t have a ticket,
you can get a ticket from the registration desk for this evening. It is wet out, for
those who haven’t seen the weather, but — and the building is closed, so it is now secure
building, so it is only out after you leave the auditorium. But I just want to thank everyone,
and a round of applause for all of our panels today. [applause]

Leave a Reply

Your email address will not be published. Required fields are marked *