- So at this point, anyone who would
like to ask a question to any and all of the speakers
should line up behind the microphone.
And can we have your name, please?
- Hi, I'm Evelyn [? Ginsberg. ?] This
is a basic genetic question.
If we share 99 point something percent of our genes
with chimpanzees, but my distant cousins on 23andMe share 2%,
is this a different counting system?
How are we counting what percentage of genes match?
- Who's going to answer that?
- I can try to speak to that since I brought up the factoid.
So 99.9% of your DNA is the same as, say, your neighbors,
or somebody across the planet.
That said, that fraction that varies
tells us a lot about inheritance.
So if you're looking between you versus one of your parents,
you will share 50% of the variable sites in your DNA
with your parents.
And then if you start to look back further and further
in time, say, look at your grandparents,
you'll share on average about a quarter of your variable sites
in your DNA with your grandparents.
So it's a difference between what's varying generally
in the population versus what varies across enormous time
scales.
- So it's the variation between different humans,
assuming that you've found all the variation that there is?
- Yeah, that's right.
And this is due to common variation, not
these super, super rare individual specific variants
that are arising very newly within the population.
- So it's some small subset of all possible genes
that you're measuring when you say we have
2% in common with Neanderthals.
- Yeah, so that refers to what fraction
of your genetic ancestry traces itself to the Neanderthals.
So the 2% number says, essentially you
can think of every point in your genome,
you can trace its ancestors.
And so you have many such genetic ancestors.
And of those set of genetic ancestors, about 2%
will trace their ancestry to the Neanderthal lineage.
So that's really what we mean when we talk
about 2% Neanderthal ancestry.
- I'm not sure I understand.
2% of the entire genome, counting every single DNA pair?
- So it's the ancestry along of your genome.
So for example, when you're looking at parents,
right, so that is genetic ancestry that's immediate.
But then when you go to the next generation,
you have your grandparents, and there
are four possible ancestors.
And they can contribute different amounts
of genetic material to different parts of your genome.
- Oh, so it's 2% of the tree?
- So the way to think about it is,
if you take one point in your genome,
and you go to your parent--
so you inherit it from one of your parents.
And then you follow which parent that parent inherited it from,
and you keep going all the way back.
So you will, at some point, maybe 50,000 years back,
you will either go to the Neanderthal,
or you will go to the modern human.
So this is for one point in your genome.
Now, you go to another point, and you repeat this process
and see which ancestor that part of the genome traces
its ancestry to.
In general, they need not be the same,
and this is because of recombination.
So it keeps switching around as you move along the genome.
And so that's where--
then you count it up, and you ask,
what percentage comes from Neanderthals?
On average, it's about 2%.
- Thank you.
- Thank you.
Next.
- Hi.
My question is for Alicia Martin.
And this problem, right, that we have, it's not a new problem.
I think of it as sampling error, where the data that we collect
and the conclusions that we draw from it in medicine
have been around for a really long time.
So this seems to me to be an extension of a very, very
old problem, and the fact that we draw conclusions
in ways which somehow don't really take into account
what the actual sample is.
Do you have any thoughts on how we
can go about correcting this?
- Yeah.
That's a great question.
Thanks.
Yeah, so to your point, this problem
has been around for a long time.
It's been around as long as genetics,
or even far before genetics was being studied regularly
across populations.
So like a decade ago, for example, researchers
said that 96% of participants in genetic studies
were of European descent.
Now we're 80%.
So we're making some progress, but of course it's
not fast enough.
And so I think the ways that we need
to start enabling some parity in these public health issues that
are going to start emerging with the translation
of genetic technologies is by enabling funding sources
to start to focus more outside of European populations
than within European populations.
I think we're also going to need to encourage
more genetic investment of resources
outside of Europe and the US.
Local researchers are going to need
to take this up at scale within different continents, as well.
I think this is going to need to happen on a very large scale
globally.
But it's also important to note that we're not
reflecting the diversity that's present in the US.
So the US is also going to need to step up and make sure
that the data access for some very large, diverse cohorts
that already exist are accessible to a globally
diverse community.
So there are some very large cohorts
that are out there that capture the breadth of diversity
in the US, where it's actually really hard to access
genetic summary statistics.
And this isn't due to any privacy concerns,
because genetic summary data is exactly that.
It's summary data.
There's no individual level privacy
concerns for that type of data.
So we also need to make sure that we can access that.
So there's many different moving parts
that all need to coalesce, I think, to move towards parity.
And it's going to be a very hard problem
that we need to start really making concerted efforts to try
to address over the coming many years.
- Just one quick follow up, and that is, do you
think it would help to diversify the professional community
of researchers doing the work?
- Yeah.
[LAUGHTER]
It feels a little ironic talking to some
of my more diverse colleagues who
are studying European participants at scale
as sort of the white girl up here harping on the fact
that we need more diverse participants
in genetic studies.
Of course that's the case.
And I think we certainly need to diversify
the academic community.
But we also need to encourage new leaders that
are invested in this problem as something
that they're going to be driving forward longer term.
So yeah, I absolutely agree with that.
- Nick?
- I'm Nick Patterson.
I used to be at Broad, and I'm at Harvard Evolutionary Biology
and Harvard Medical School.
And I just wanted to share with you something I talked about
with Alicia, how some of the minority
communities in this country have not been treated well,
and there's a second order effect on whether they
participate in studies.
So I just will share with you, I was
involved in a medical study of cardiovascular genetics
in African-Americans.
And the primary center of the study was in Mississippi.
And we had a poorer response than we
wanted from the African-American community.
And a common question we were asked
is, if we give you our genetic data,
will that be available to law enforcement?
And the legal answer is, under subpoena, absolutely yes.
And it's hardly a secret that the African-American community
in the deep south and law enforcement
have not always got on very well.
So the consequence was that our participation
was lower than we wanted.
And so that's a political problem
due to past political and present political effects
in the United States.
And it's causing technical difficulties
in acquiring the data that Alicia wants to see.
So I just thought I'd share that.
- Yeah, Nick, that's an excellent point.
Thank you so much for bringing that up.
I forgot to mention that, absolutely, the mistrust that
happens in academic research is often
earned through historical misuse of data
or through historical tensions with law enforcement.
That's absolutely the case.
And I think to start to broach some of these issues we need
to be very delicate about how we start to communicate both
the needs for public health purposes versus the tensions
with law enforcement or with other bodies.
- Yes, Alyssa?
- Hi.
So I have my freshman seminar students here today,
and they're taking a freshman seminar about prediction.
And some people here know I have a particular interest
in prediction.
And my long term dream is that data science will help us
with climate change.
And so what does this have to do with you?
The question is, I know that in academia there
are a lot of silos.
And a lot of what you are doing is obviously related
to what each other are doing here, right?
And what I want to know is, let's say
we're thinking about a future where we're
going to have the ability to really simulate
the future of the earth.
And we need to understand the interaction of all four things
that you were talking about.
And how is the climate forcing that?
And how are the fact that people are eating tilapia, and ones
with Neanderthal DNA are, and [INAUDIBLE],,
and then they're having clover honey on it, and whatever else
they're doing.
So what venues are there for you to exchange
your data and your ideas, other than today, which is wonderful.
But in other words, are there organizations
where people are thinking about pooling the kind of data
and research that you all have across your sub-disciplines
that all touch on evolution?
- Anybody can answer.
- Anybody answer that.
- I mean, I don't know about the others,
but I'm a strong proponent of open science.
I put all of my data sets, all of my code,
drafts of my manuscript, everything is just available
online even before publication.
Which might be a pretty extreme route.
But I think, I believe in just having
it out there for other people to use and see.
Now, I know that's not always the case,
especially if you're dealing in human genetics
and you have samples that maybe would violate some sort of law.
And you can't just put it out on the internet for everybody
to have access to.
But at least nobody cares about clover to that extent.
So I'm happy to put all of my data and my code online,
and it's often helpful for other people,
too, who are doing similar things.
So at least in my case, I just try to be as open as possible,
as I can, with the stuff that I am doing and want to do,
and then hopefully that fosters that sort
of open science type of idea that you're talking about.
- And that, if I may, and that is the problem right there,
a big portion of the problem is that, the white clover,
who really cares?
But the implications for humans, and of course tilapia
aside, is that we are restricted by putting human data
or identified-- everything is de-identified.
And it's not, from my experience,
it's not so much that the participants are worried
about what they put online.
It's really because of the health insurance
discrimination.
There's other-- we have to go deep,
and it's going to be more than the two minutes that
are left in this session, to uncover
why we can't have open access.
My experience is not because there's not a willingness
to be transparent and forthcoming
by the participants, but they're hindered by forces
beyond themselves.
So I'm going to open it up to the rest.
I didn't mean to get on my soapbox.
But go on.
Anybody else?
Yes, Katie.
- Just one other note.
I think that that idea requires us to really think
outside the box.
It's not only data accessibility,
it's thinking about how we cross disciplinary boundaries to make
those kinds of connections.
And I think that's a persistent societal challenge.
I don't think that there is an organization that does that,
but it requires individual thinkers to be
willing to take those leaps.
- Yes, sir.
A few questions?
- Hi.
OK, I have a couple questions.
Number one, in your talk, you mentioned
that there's 2% Neanderthal DNA on average in a person.
And I noticed in that graph where
you had the figure where you had the amount of introgression
in, I think, East Asians and some other populations,
the peaks were not overlapping.
So I'm just curious, what's the total coverage
of Neanderthal peaks across everyone?
So if I have 2-- well, I don't have 2%.
But if someone else has 2%, and someone else has 2%,
that might be 3% total coverage.
Are they very overlapping, or just completely everywhere?
- Yeah, so we tried to do this experiment where we said,
let's say we took all of the segments of Neanderthal DNA
across all the people that we can identify.
Can we stitch them together?
And if you do that, how much of the Neanderthal
do we actually recover from present day people?
And so we recovered of the order of around 60% of the genome.
Yeah.
So-- six-zero, in terms of length.
So what that's telling you is, at least one person has
a chunk of Neanderthal DNA that you can use to stitch together
a full kind of genome.
And now you can ask, I mean, there's
another question you can ask, maybe how much further can we
get?
If we had a bigger sample size, can we get closer?
And I think that's where you run into these issues
of these deserts, which nobody seems
to have Neanderthal DNA at.
And so I think they're kind of saturating, literally,
how much of their genome we can cover that way.
- OK, well, that's already a surprise.
I thought it would be like 8%.
Is that a published analysis, or is
that just like some experiment?
- It's published in our 2014 paper.
- OK, sweet.
All right, second question, because I was confused.
You had this argument of constraints to justify--
actually, I completely forgot.
I just want you to re-explain how constraint on evolution
helps you make arguments for introgression events, sort
of re-explain.
- So I'm assuming you're talking about what happens in regions
of the genome that are constrained,
and how much Neanderthal DNA is present in those regions?
- Yeah.
You just had some graphs that's like constraints on the x-axis.
I just didn't understand that.
- Yeah, so there are some regions of the genome,
for example, if you go close to genes, which are functional
elements in the genome, that tends to be more constrained.
So what that means is--
- No, yeah.
Yeah, I know what that means.
I'm just saying, how does that play into the introgression?
- OK, so those are places where there
is going to be stronger selection on mutations
that are deleterious.
So for example, if you have a Neanderthal chunk that
carries a deleterious mutation, depending
on which part of the genome you're looking at,
whether it's a more constrained or a less constrained genome,
the selection is going to be stronger
in one versus the other.
And so in the first case, it gets removed quickly.
And so you're going to see a reduction compared
to the second case.
- OK, cool.
And then, sorry, one more question for Alicia.
- Can you save your question for the reception?
We'll have a reception downstairs right after,
and then you can corner [INAUDIBLE]
and ask all the questions you want.
I'm sorry.
If your question is quick?
- Yeah, really quick.
My name's Christina Wilson, and I'm
thrilled by the advances in research
that you all are doing.
But I also wonder what you personally
feel like your obligations are in a country like America
where, for instance, health care is not
a guaranteed-- access to health care is not a guaranteed right.
And what the implications of essentially marking everyone's
pre-existing conditions from birth, I think you said,
has on the sociopolitical economic conditions that
may or may not be exacerbated.
One thing, to the organizers, I would have loved
to see a bioethicist or a--
- We talked about that.
Sean, didn't you say during the break that-- yes,
we've already taken that--
- Oh, great.
So that's fantastic.
But also personally, I mean, we already
heard from James that his philosophy
is put it all out there.
And you're dealing with clover.
But this happens in cities, too, with open data in cities,
about technological points of contact and similar things.
You have corporations and for-profit industries
using this public data to do what they will with.
So I'm just wondering what, as scientists,
what your personal sense of obligation
is to protecting the rest of us living--
and being thrilled by the discoveries,
but wanting to live a good life.
- Thank you.
Yeah, that was a pretty big question.
So I think maybe we can start to tag-team pieces of this.
So in thinking about how genetics can mark you
from birth for some traits or not,
I want to re-emphasize the fact that genetics is not
deterministic, especially for complex traits
that we're studying like cardiovascular disease.
Of course, your environmental effects
can be as big or more important than some
of your genetic factors.
So I think it's, of course, an important issue
to make sure that the non-discrimination acts
in genetics continue to be upheld.
We need to make sure that that does not
go away at any point in time.
I think we need to advocate to policymakers
that this is really important as it stays separate from health
insurance companies.
All of those factors are really important.
And then, to the open science sense of camaraderie
that we all share, I would emphasize
that, in human genetics, while it's
hard to share individual-level data,
we are also interested in sharing summary data that
can be used for anybody that's out there to compute
their own scores or start to learn about their own biology.
So that's certainly also out there
and accessible through a lot of these public servers.
And we also put all of our code out there,
and we also put all of our manuscripts
on pre-print servers, so that anyone can access them
as soon as we're in a state to actually share them.
Yeah.
Does anyone else want to add anything?
- Yeah, so I'm also part of the same community,
and so again, I think it's extremely useful
to have access to data.
So in human genetics, a lot of the genotypes,
including the Neanderthal DNA, that's publicly available.
You can access them.
You can search for them.
You can write your own tools to query your DNA.
Harder when you overlay it with free data of different kinds,
but that's where the community has agreed
to the sharing of some of these statistics,
and that's been a way of still sharing data without getting
into these privacy issues.
And I think that's been extremely valuable.
- So again, [INAUDIBLE] that you're talking about
[INAUDIBLE] which is great, talking about the bottom-up
effect of us taking that knowledge that we now know
about our personal genome, like taking a 23andMe or whatever,
and making changes in our lifestyle.
But what I'm talking about is a top-down.
And so being able, at a policy level, or a corporate level,
to take that data and apply it to mass populations.
[INAUDIBLE]
- OK, well, that is great for reception discussion.
I would like to thank wholeheartedly these four
fantastic rock stars.
[APPLAUSE]
Sriram, Alicia, Katie, and James.
Thank you.
I am very optimistic.
The future is very bright.
Because these are early curious scientists that will inform us
for many, many years to come, years that Alyssa
and I will no longer be here--
no.
Didn't you take your 23andMe?
I'm kidding.
OK, with that, the reception is on the first floor.
Thank you.
[MUSIC PLAYING]
For more infomation >> LCFC Professor in Child Health - University of Leicester Royal Visit 2018 - Duration: 0:50. 
For more infomation >> University cleaning dorms after mold reports - Duration: 0:25. 
For more infomation >> University cleaning dorms after mold reports from students - Duration: 2:40. 
For more infomation >> Old Dominion University students making gifts for kids in need - Duration: 1:40. 
For more infomation >> Clovis High School Chamber Choir Performs at West Texas A&M University - Duration: 0:21. 
For more infomation >> Heartwize Project - University of Leicester Royal Visit 2018 - Duration: 1:09. 
For more infomation >> Ca. University Used MKUltra To Make Students Pro-Immigration - Duration: 5:00. 
For more infomation >> Practical Theology Doctorate at the University of Birmingham - Duration: 5:16. 
Không có nhận xét nào:
Đăng nhận xét