[00:00.000 --> 00:05.600]  Welcome, everybody, to our panel discussion, Biomedical Data, Open, Secure, or Stolen?
[00:06.140 --> 00:17.220]  We are so excited to talk to you today about open data and open science and the way that they're currently championed in academic and the public sector to promote innovation and discovery.
[00:17.240 --> 00:21.860]  However, not all data can be open, and the reasons are wide-ranging.
[00:21.980 --> 00:26.200]  And I'm so excited to be joined by my colleagues to start this discussion.
[00:26.260 --> 00:27.880]  So my name is Michelle Holko.
[00:27.880 --> 00:29.380]  I currently work at Google.
[00:29.380 --> 00:36.160]  My background is I have a PhD in genetics, and I've worked a lot in bioinformatics, both for cancer and infectious diseases.
[00:36.180 --> 00:39.640]  But the majority of my career I actually spent in the public sector.
[00:39.640 --> 00:58.880]  I worked at NIH for a while, and most recently worked in the White House as a Presidential Innovation Fellow, and worked on a wide variety of approaches to science, really at the intersection of how technology can enable science and discovery, but that also that there are unique security aspects associated with that.
[00:59.320 --> 01:02.640]  Crystal, I'd love to turn it over to you to introduce yourself.
[01:03.140 --> 01:04.340]  Thank you, everyone.
[01:04.340 --> 01:06.520]  Hello, I am Dr.
[01:06.520 --> 01:07.480]  Crystal Sosi.
[01:07.480 --> 01:13.200]  I have a PhD in genomics and health disparities from Vanderbilt University.
[01:13.200 --> 01:18.560]  I'm transitioning to an assistant professor of faculty role at Arizona State University.
[01:18.560 --> 01:24.200]  I'm also with Joseph Irscheda, a co-founder of the Native Biodata Consortium.
[01:24.200 --> 01:47.020]  My background is largely population genetics and also bioethics, and as an Indigenous scientist, my perspective is that genomic data, particularly in health data, should belong to the communities from whom it comes from, and therefore we should be thinking about benefits of data and data science for the community's perspectives,
[01:47.020 --> 01:51.480]  not necessarily researchers' right, supposed right to access to data.
[01:51.480 --> 01:52.340]  Thank you.
[01:52.340 --> 01:55.820]  I'm going to turn it over to, I guess, Joseph Irscheda.
[01:58.060 --> 02:00.520]  Hi, I'm Crystal.
[02:00.520 --> 02:04.300]  Like she said, we're the co-founders of the Native Biodata Consortium.
[02:04.300 --> 02:08.160]  I'm the vice president and executive director here.
[02:08.160 --> 02:30.000]  I am of Mexican descent, but the reason that I'm here on the Cheyenne River Sioux Reservation is because my wife and kids are enrolled here, and these questions about Indigenous data sovereignty are important to me because of that relationship with the Cheyenne River Sioux Tribe,
[02:30.000 --> 02:42.700]  but also because a lot of this data is unprotected for urban Natives, for Natives who were formerly recognized but were terminated by U.S.
[02:42.700 --> 02:55.700]  government, which ended in 1975, and then all of the Indigenous people in Latin America who are not protected that live in this country and are being sampled in those countries for the United States.
[02:56.600 --> 03:07.220]  But I have a master's of pharmacogenomics from the UW-Seattle, and I'm just finishing up my doctorate at Johns Hopkins for environmental health and engineering.
[03:07.220 --> 03:08.160]  Thank you.
[03:09.200 --> 03:09.920]  Awesome.
[03:09.920 --> 03:10.360]  Larry?
[03:10.360 --> 03:11.320]  It's me, yep.
[03:11.320 --> 03:17.260]  Hi, I'm Larry Hunter, and I am a long-time advocate of open science.
[03:17.260 --> 03:22.940]  I'm a professor of biomedical informatics at the University of Colorado School of Medicine.
[03:22.940 --> 03:27.920]  I was the founding president of the International Society for Computational Biology.
[03:27.920 --> 03:44.200]  I have a very long -standing interest, more than 30 years now, trying to change scientific publishing so that scientific papers are more accessible to more of the world, to build capacity within communities that have not traditionally been well served by science,
[03:44.200 --> 03:54.260]  to do science on their own, and in general to try and make science work better for the people of the world.
[03:54.580 --> 03:57.120]  And I'm also an active scientist.
[03:57.120 --> 04:06.940]  I've published a couple hundred papers in machine learning and artificial intelligence and natural language processing as related to biomedical applications.
[04:07.840 --> 04:09.300]  Awesome, thank you.
[04:09.300 --> 04:26.680]  Yeah, so, you know, one of the things that I think was kind of a precipitating thought to this topic and this discussion is the fact that there's been a huge emphasis on equity, you know, especially in light of the COVID pandemic, but even, you know, much longer than that,
[04:26.680 --> 04:38.240]  going back historically around, you know, who gets to do science, who benefits from science, and how do we actually do that in a way that makes things a lot more equitable?
[04:38.240 --> 04:50.200]  And I think, you know, that's one of the things that we've talked about at length in terms of, you know, openness can actually often lead to expanding who's able to participate.
[04:50.300 --> 05:04.920]  But then how do we ensure that people are benefiting, people that are participating in science as either research participants or as scientists themselves, how they're able to bring back that benefit to themselves and also their communities?
[05:08.680 --> 05:10.760]  Can I push back a little bit on that?
[05:10.760 --> 05:11.140]  Yeah.
[05:11.140 --> 05:16.220]  Because I actually didn't hear enough conversations about data equity.
[05:16.220 --> 05:22.720]  I heard more conversations about the need to include diverse and underrepresented peoples in science.
[05:22.720 --> 05:28.740]  And so we have the diversity part and the inclusion part, but I didn't hear enough of the equity, actually.
[05:29.020 --> 05:30.760]  And to me, that was disconcerting.
[05:30.760 --> 05:50.000]  Because if we think about it, if we only emphasize the D and I as part of the EDI efforts or DEI efforts, and ignoring equity, we have, unfortunately, we run the risk of sort of repeating some of the ethical controversies that have occurred in biomedical research in the past.
[05:50.000 --> 06:02.300]  So I mean, it's astonishing to me in the field of human genetics, and also in the next generation of data scientists that enough of them are not being trained in the histories of their own fields.
[06:02.300 --> 06:21.880]  And I'm not talking about like, several decades ago, I'm talking about even just one to two decades ago, such that it's whenever I lecture in academic institutions worldwide, I always feel like I have to constantly be re-educating or educating people about the Human Genome Diversity Project,
[06:21.880 --> 06:38.780]  and 1000 Genomes Project, and also the Genographic Project, which were large scale diversity projects of the 90s and early 2000s, that were denounced by hundreds of indigenous groups worldwide, because of concerns related to co-optation and commercialization,
[06:38.780 --> 06:55.240]  and patenting of their genomic information, biological information, especially since it is culturally incongruent with how they viewed whether or not these things should even be owned by, you know, entities.
[06:55.260 --> 07:14.840]  And, you know, the first people to benefit from openly available indigenous biomarkers, for instance, were not the communities who are still dying and waiting for medicines and therapeutics to arrive, as promised, but rather it was for-profit companies,
[07:14.840 --> 07:21.860]  like direct-to-consumer genetic ancestry companies, like 23andMe and Ancestry, that profited billions of dollars.
[07:21.920 --> 07:42.300]  And then, so, you know, when we sort of talk about where we are now in the 2020s, I'm still seeing in these new and emerging and existing large scale diversity projects, this rush to include diverse and underrepresented peoples, but I'm not hearing, again,
[07:42.300 --> 07:44.400]  that concern about equity.
[07:44.660 --> 08:00.380]  So, I mean, I really would love to hear more about it, but I keep hearing what we keep saying, like democratizing data, and I hate, I can't tell you how much I hate that buzz phrase, because, I'm sorry, what model of democracy are we modeling here?
[08:00.760 --> 08:04.660]  We're talking about the garbage fire of the American democracy system.
[08:04.660 --> 08:09.880]  I don't think anybody should be modeling anything about, with that system.
[08:09.880 --> 08:23.980]  And democracy generally means a majority rule system are, and in that system, the smaller underrepresented groups are continually being disenfranchised.
[08:23.980 --> 08:34.120]  So, for me, democracy or democratizing is not the same as equity, and I really hate it when people confuse the two or don't have enough access to equity.
[08:35.460 --> 08:53.780]  Yeah, I totally agree with that point, and, you know, one of the, there are definitely great examples in genomics, but more and more, what we're seeing is, as technology like wearables are starting to be highly ubiquitous, there's still, again, this equity piece that's missing,
[08:53.780 --> 09:03.960]  where not all groups have access to them, the algorithms are being developed on certain types of groups, and then yet, who has access to the data?
[09:04.300 --> 09:11.580]  You know, I can't even get the raw signal for my data, I get the whatever process version that they want to give me.
[09:11.820 --> 09:25.540]  And so, how do you, you know, in a way, I feel like the innovation has to come from within, as well as, you know, from companies and other organizations.
[09:25.540 --> 09:41.440]  And, you know, I think that the work that you guys are doing with Native Biodata is so important in this, because even just creating awareness in communities about how your data are being used, how you can take control of your data, and how you can use your data for good.
[09:41.520 --> 09:53.880]  But at the same time, I think you're spot on that the equity piece, there's been a lot of talk about it, but I haven't really seen a lot of great innovation around equalizing value.
[09:54.500 --> 10:02.540]  Well, this is one of those places where I want to, you know, make claims that more openness about data would be better for everyone.
[10:02.580 --> 10:04.160]  So, wearables is a great example.
[10:04.160 --> 10:11.940]  It is very hard to get access to wearables data, even though lots and lots of people are walking around with Fitbits and Apple Watches and the equivalent.
[10:12.320 --> 10:19.440]  And we've had a recent push to have a right to repair where you're allowed to open your thing and, you know, you bought it, you can work on it.
[10:19.440 --> 10:21.660]  We don't have a right to your own data.
[10:21.660 --> 10:39.020]  And so the data that people are generating about themselves, really detailed, deep, personal data, biomedical data in some degree, they don't have access to and isn't widely shared and is used mostly for the benefit of the shareholders, the companies that invented this stuff.
[10:39.020 --> 10:47.320]  And often that means for selling ads or other maybe less than, you know, socially beneficial activities.
[10:47.320 --> 10:59.880]  Whereas if that data were more freely available, then it's possible that creative folks, because it is possible to do science now with the kind of technology that most of us have access to, right?
[10:59.880 --> 11:01.720]  Computers and the internet and the like.
[11:01.720 --> 11:10.500]  You could do a lot with access to the right data streams and maybe people could do things that would be more socially beneficial if the data were more available.
[11:10.500 --> 11:16.380]  Even if it were just, you know, by virtue of individuals saying, I want my own data and then contributing it to something.
[11:17.160 --> 11:27.120]  So this is, in my view, this has been a product of a slow and explorable push towards privatization starting just before the Reagan era.
[11:28.220 --> 11:36.380]  Most of research has been done with publicly available tax funds, but even then that data ends up being privatized.
[11:36.380 --> 11:44.900]  So the reduction of the commons has been something that's been happening for a long time, including like water rights and water resources, right?
[11:44.900 --> 11:49.280]  Almost every state now bans people from collecting rainwater.
[11:49.480 --> 11:52.300]  So this is something that's been happening for a long time.
[11:52.300 --> 12:05.180]  And so the commons used to be open and things that you cited as having open data and open access should be available to everybody, but it's not because of capacity and capability.
[12:05.180 --> 12:14.080]  And what I mean by capacity and capability is you can give me a Lamborghini, but if I'm five years old, I can't drive that Lamborghini.
[12:14.080 --> 12:21.060]  Or if you give me a Lamborghini and I don't have a road that's built so I can drive the Lamborghini, that Lamborghini doesn't do me anything.
[12:21.060 --> 12:29.560]  And so one of the advantages of the commons, at least prior to privatization, was that I could build capacity in that way.
[12:30.880 --> 12:43.200]  There's no longer an avenue for me to do that because as you said, these big companies control the data and they also control the ability to get funding to use the data, right?
[12:43.200 --> 12:53.380]  So we were just hearing not too long ago that some of these companies were trying to do R&D by themselves, but it's too expensive for them to do.
[12:53.380 --> 13:10.740]  So they're almost reliant on the NIH and the NSF to create these data sets for them, and then they take it the rest of the way with their ability to get the licenses and get the approvals from the United States to make the product and safety test the product,
[13:10.740 --> 13:13.460]  which the rest of us don't have.
[13:13.720 --> 13:25.580]  So it's just, it's been a slow process to take away much of what belongs to all of us and put it into the hands of the few that can go through the system.
[13:29.700 --> 13:49.500]  Yeah, and I think we've seen a lot of examples, especially in the biomedical research space, where people participate in research or companies are doing research or in the academic space without a really clear return of benefit.
[13:50.140 --> 14:08.520]  And the consent process, I've been through it several times, and there's actually one study that I'm still trying to get unblinded from that happened, gosh, about how old is my youngest son five years ago, and they still haven't told me if I got placebo or the active compound,
[14:08.520 --> 14:10.540]  right, to try and infect him.
[14:10.880 --> 14:24.440]  And I think that there's a real issue with giving people access to their data, to their information, and also giving them a say in what happens with it.
[14:24.440 --> 14:41.420]  You know, I think that this is one of the areas where I see there's a huge opportunity for technology to help us with this, but only if we have a really good understanding about what types of options people want and that are going to feel most supportive,
[14:41.420 --> 14:43.720]  both to individuals and communities.
[14:45.540 --> 14:50.000]  Well, and Joe, I don't want to argue with you about the closing down of the comments.
[14:50.000 --> 14:55.160]  I think that, you know, rampant commercialization of everything is clearly a fact of the modern world.
[14:55.160 --> 15:01.080]  But the olden days were not also great, at least around access to scientific data and science.
[15:01.260 --> 15:13.220]  And so it used to be the case that if you couldn't afford a very expensive subscription to the journal Nature or Science or Cell, and you didn't have a good library nearby, you were kind of out of luck.
[15:13.780 --> 15:18.180]  You couldn't even read about other people's science, let alone do your own.
[15:18.560 --> 15:31.920]  And I think one thing that technology has given us is the ability to shine a light on scientific publishers as basically profiting in ways that, you know, we're not providing any benefit.
[15:31.920 --> 15:35.520]  There were no longer printing presses and postage to pay for this stuff.
[15:35.520 --> 15:37.620]  Anybody could download it off the internet.
[15:37.800 --> 15:42.380]  And so those gatekeepers who made it so expensive made it hard to do science.
[15:42.380 --> 15:47.480]  And in the last 25, 30 years, it has really been a revolution in the open access to science.
[15:47.480 --> 15:52.820]  So it's now much more possible for anybody to read existing science.
[15:53.080 --> 16:14.500]  And given that so much science is now being done in silico, that is by computers and analyzing datasets, often public datasets, it becomes more accessible to people who want to do something, who have some particular scientific quest, a question they want to answer that matters to them or their community that hasn't previously mattered to the powers that be.
[16:14.500 --> 16:21.940]  So I think there's some good things happening as well as this, you know, commercialization of everything.
[16:22.960 --> 16:26.240]  Again, I don't argue with you either, but I think we're close to the same age.
[16:26.240 --> 16:26.900]  I'm 56.
[16:26.900 --> 16:30.820]  So I did research when we didn't have high throughput computing.
[16:30.860 --> 16:47.960]  And much of the publicly funded research was very siloed, like even among department members that were working on similar things, because mostly their notoriety, not so much their economic benefit, was dependent on whether or not they could publish something that was unique,
[16:47.960 --> 16:48.700]  right?
[16:48.700 --> 16:51.160]  And nobody cared back then.
[16:51.160 --> 16:58.240]  It's only because now we have the computing capability to share this information widely that they're demanding open access.
[16:58.240 --> 17:10.040]  I mean, the things that spurred that on, you were probably aware at the time in the 90s when graduate students were shooting their advisors because they didn't graduate or numbers were fudged.
[17:10.040 --> 17:14.940]  And there the data was siloed, right?
[17:14.940 --> 17:20.680]  So you want to make the open data or the data open so that those kinds of things don't happen.
[17:20.680 --> 17:27.860]  A graduate student has recourse to talk about their advisor, you know, and say, this is what I produced.
[17:28.020 --> 17:33.260]  Other people that want to look at it to make sure that your calculations are correct and it's reproducible.
[17:33.260 --> 17:35.000]  Those are all good things.
[17:35.000 --> 17:53.680]  But the open data in general, I think allows for people who don't have the capacity, like even small university researchers who don't have the capacity can't go as quickly as these universities with these huge computing centers, these huge genome sequencing centers.
[17:53.680 --> 17:59.810]  So now they're at a disadvantage , not just native people, but even the small universities that are disadvantaged.
[18:03.160 --> 18:11.760]  I think one of the important things to realize is that, first of all, not all data has the same amount of risks, right?
[18:12.000 --> 18:33.420]  As much as I'd like to have parallels with, for instance, open access publishing, the risks of a paper being private is going to be or not is different from a person's own genomic information, which has the ability to reidentify not only themselves, but everyone they're biologically related to.
[18:33.420 --> 18:54.130]  And that risk is even more disproportionate if you're talking about a member of a small racially identifiable population, which you can open up to negative stereotyping and concerns, for instance, like genomic or other forms of racial profiling.
[18:54.890 --> 19:01.910]  So that's one thing that we need to understand here is like not all data has the same amount of risks.
[19:01.910 --> 19:09.510]  But then also, not all data is beneficial to everyone in the same way, and particularly in terms of communities.
[19:09.630 --> 19:26.630]  And just hearing the back and forth between Larry and Joe, there's a lot of microcosms, for instance, in paleogenomics research or ancient DNA research, that in ancient DNA, first of all, there aren't that many human ancestors available.
[19:26.630 --> 19:42.710]  So it feels like our indigenous human ancestors particularly are treated like gold, because there's such a finite number of quality DNA samples from indigenous ancestors.
[19:42.710 --> 19:58.550]  And there's a concern that researchers and museums are not disclosing who they have in their bio collections, because they don't want to be forced to return or repatriate those ancestors back to descended communities.
[19:58.550 --> 20:09.690]  But at the same time, researchers, ancient DNA researchers, they both want to sort of keep their cards close to their chest, so to speak, and not revealing what they have on hand.
[20:09.690 --> 20:17.270]  But at the same time, also asking for access for other people to reveal what they have within their collection.
[20:17.570 --> 20:27.970]  There are only a few number of ancient DNA labs in the world that have the capacity to do this type of work that they do, and they publish in high-impact journals.
[20:28.010 --> 20:34.510]  And a number of them have best-selling New York Times books.
[20:34.510 --> 20:39.630]  So they not only benefit professionally, but also commercially.
[20:40.770 --> 20:44.370]  And, you know, obviously those universities also benefit.
[20:44.370 --> 21:01.030]  But what nobody seems to... or what everyone seems to forget, who are so in the weeds of their academic disciplines, is that population migration research largely has no interest or benefit to indigenous research communities, right?
[21:01.030 --> 21:14.550]  So just building that back out, you know, we really have to be questioning, you know, who are the actors that have the ability to first benefit?
[21:14.550 --> 21:27.410]  And are those entities also interested in ensuring that the data is equitable to everyone else, including their competitors, beyond themselves, right?
[21:27.410 --> 21:34.010]  Because not everyone has the benefit, you know, wants to act for the benefit of society.
[21:34.010 --> 21:50.990]  So that's kind of the trouble with some of these models that want to, of course, encourage social benefit, that there are very powerful agents that don't care about social agency or benefit, and they only have their own concerns at hand.
[21:52.590 --> 21:54.470]  Yeah, that's a really great point, Crystal.
[21:54.470 --> 22:03.390]  And I think, you know, even if people do have intentions that are good, they still make mistakes, right?
[22:03.390 --> 22:16.530]  I mean, I think that that's one thing is that a lot of, a lot of times you assume that, you know, what your level of comfort of data sharing is, you know, shared amongst others.
[22:16.530 --> 22:19.590]  And that's something I think that we really can't take for granted.
[22:19.830 --> 22:31.710]  And I think that we really need to be a lot more thoughtful and intentional about, you know, that whole, not only the consent process, but the return of information, as well as, you know, the economic models and benefits there.
[22:31.710 --> 22:42.230]  And I think that, you know, there's a need for, you know, policy and technology innovations, but also, you know, thoughtful approaches to addressing the economic issue.
[22:43.050 --> 22:44.030]  Michelle, I'm shocked.
[22:44.030 --> 22:46.770]  I thought you were going to raise the national security stuff, right?
[22:46.890 --> 22:48.630]  You're our defense geek.
[22:48.830 --> 22:56.990]  Arthur, you know, when Crystal says there are bad actors out there who might do bad things, I was ready for you to jump in and say, yeah, we have to protect everybody.
[22:56.990 --> 23:00.710]  So what do you think about the national security pieces of all this?
[23:00.890 --> 23:03.090]  Yeah, I mean, well, it's definitely there.
[23:03.210 --> 23:12.090]  I also think, though, that, you know, security is the kind of thing that you're never going to get it perfect.
[23:12.270 --> 23:17.950]  And you also have to weigh the risks and the benefits, right?
[23:17.950 --> 23:26.950]  And I think that in most cases, it is beneficial to share information, you know, to promote discovery.
[23:27.070 --> 23:39.130]  However, I do think that, you know, especially for data collection efforts that are of national security significance, yes, for sure, those need to be protected and should not be open access.
[23:39.430 --> 23:46.070]  And so, you know, building technology systems that can achieve all of that, I think is critical.
[23:46.070 --> 23:48.430]  And that work is kind of ongoing.
[23:49.090 --> 23:55.870]  But there's also, I think, a need for understanding and bringing back.
[23:55.870 --> 24:01.590]  So there's this kind of fear that happens in some communities around citizen science.
[24:01.590 --> 24:09.930]  And the fact that, you know, desktop oligo sequencing, or creation of oligos and synthetic biology.
[24:10.130 --> 24:12.430]  And, you know, now everybody can do that.
[24:12.430 --> 24:18.090]  And, you know, they're going to be rogue people, you know, building monkeypox and unleashing them on the world.
[24:18.090 --> 24:22.470]  And, you know, when it comes right down to it, it's technically not that easy.
[24:23.450 --> 24:26.590]  You know, there are some technical hurdles to those types of things.
[24:26.590 --> 24:34.070]  But I really think that, you know, overall education and understanding is really the way to go.
[24:34.070 --> 24:45.510]  You know, there's also technical challenges around, you know, say you have engineered the bio threat agent of interest, there are dissemination challenges with that.
[24:45.510 --> 24:50.630]  And so, you know, yeah, we've seen infectious diseases are actually a pretty good one.
[24:51.070 --> 24:55.710]  But it depends on the infectious disease, it depends on a lot of factors.
[24:55.710 --> 25:04.730]  So, you know, I think that it's, there are definitely scenarios where I don't think that it's appropriate to have open data and open sharing.
[25:05.250 --> 25:22.710]  But, you know, another aspect of it, and then I'm going to pause, is that I think that we also need to be aware of whenever we have these types of data common systems, where, you know, data are open and freely available to the research community, the worldwide global research community,
[25:22.710 --> 25:27.730]  I think it's important to understand who are accessing those data, and for what.
[25:27.730 --> 25:34.510]  And, you know, right now, a lot of those systems are not built in a way that makes it really easy to audit that information.
[25:34.930 --> 25:49.670]  And I think that that's, you know, something that we should definitely do to strengthen our security posture, but also just our general awareness of who is participating in science, with an eye towards, you know, making sure that those are, you know, kind of on the good intention side,
[25:49.670 --> 25:57.490]  but also with an eye towards where are we seeing gaps in access to specific types of groups of people.
[25:58.300 --> 26:02.970]  I wanted to add on to what Crystal was saying in terms of context.
[26:03.170 --> 26:14.890]  So vulnerable populations or populations that are in the tails are often lumped together in big data sets inappropriately.
[26:14.930 --> 26:24.590]  And a lot of the research has to be context driven, and the statistics also have to be modified for small data.
[26:25.710 --> 26:26.870]  And I have an example.
[26:26.870 --> 26:46.830]  So we were doing an asthma study, and because of the way funding works, where you have to sort of have evidence-based modeling or evidence-based rationales for doing the study, it's kind of a chicken and the egg for a lot of Native people where the evidence isn't in the literature,
[26:46.830 --> 26:51.210]  but you have to have the evidence to get the money to do this study.
[26:51.210 --> 26:56.410]  So a lot of times Native people are forced into using a model that's not applicable.
[26:56.430 --> 27:12.590]  So we did an asthma study, and because a lot of the on the east coast, we had to use the allergen analytes for the east coast, which were a form of mold that grows on the east coast.
[27:12.590 --> 27:17.910]  The mold doesn't grow west of the Mississippi, so that's inappropriate.
[27:17.910 --> 27:27.790]  The dust mites are both genetically different and in a lot lower frequency because it's too dry for them to really have a large population.
[27:27.790 --> 27:37.110]  The analytes were developed over, you know, probably three decades on the east coast, but most of those analytes don't apply to Native populations.
[27:37.110 --> 27:55.770]  And so when we did the study, we got almost worthless data because, you know, the analytes didn't match, there wasn't money or data accumulated before, and almost no publicly funded research will let you start from scratch in a context like that, right?
[27:55.770 --> 28:02.690]  So there's a lot of different contextual issues with small population data that aren't addressed.
[28:03.130 --> 28:24.830]  And in addition, the science in general has been affected by this open data atmosphere because the Bayh-Dole Act and the America Competes Act have almost completely directed research towards commercially applicable research.
[28:24.830 --> 28:29.650]  And there's a lot of articles that are talking about how that's a bad thing, right?
[28:29.650 --> 28:35.250]  Because not all questions, at least when you start out, have commercial applications.
[28:35.250 --> 28:48.570]  And so a combination of that funding expectation and a combination of evidence-based proposals really hurts the science of doing research in small populations.
[28:50.550 --> 28:52.230]  Yeah, that's a great point.
[28:52.230 --> 29:00.650]  I mean, there's so much buzz around big data and leveraging big data for artificial intelligence and machine learning.
[29:00.770 --> 29:05.950]  But again, you have to have enough data to make that feasible.
[29:06.490 --> 29:09.670]  And so, yeah, it's a really important concern.
[29:09.670 --> 29:16.190]  But at the same time, for communities, you don't want to neglect them and say, oh, well, we're just not able to do that research.
[29:16.190 --> 29:20.070]  And then, yeah, super interesting.
[29:20.070 --> 29:32.810]  Big benefits of open science and open access, both to scientific publications and to as much data as we can share in a way that doesn't harm folks, because it does allow citizen scientists.
[29:33.930 --> 29:38.170]  You know, it does not take a supercomputer to do AI research these days.
[29:38.170 --> 29:41.270]  You can really do quite a bit with what's publicly available.
[29:41.270 --> 29:45.870]  And so it opens the door to asking different questions that matter to different communities.
[29:45.870 --> 29:57.910]  Crystal said early on that one of the things that all this ancestry tracing may or may not be interesting to some folks, but it hasn't had much benefit to indigenous populations.
[29:57.910 --> 29:59.410]  And that seems clearly true.
[29:59.410 --> 30:05.790]  So indigenous populations should be able to have the capacity to do science that matters to them.
[30:06.150 --> 30:08.230]  And I wish that were the case.
[30:08.230 --> 30:20.250]  So last summer, the NIH released two large calls for AI-driven research related to ameliorating health disparities in our communities.
[30:20.250 --> 30:22.050]  There's Aim Ahead and Bridge to AI.
[30:22.050 --> 30:29.270]  When that happened, the Native Biodata Consortium was inundated with requests to collaborate on these calls.
[30:29.270 --> 30:32.570]  But these calls were really cloaked.
[30:32.570 --> 30:41.710]  It felt like indigenous peoples being included on this grant was just a bare minimum checkbox for being inclusive.
[30:41.710 --> 30:46.010]  But in reality, they were throwing us like the smallest bone possible.
[30:46.230 --> 30:49.730]  It's like, well, you know, we want your people's data.
[30:49.950 --> 30:58.870]  And if you're not willing to give us your data, well, I guess we can throw you the bone of education engagement, but not funded.
[30:59.370 --> 31:02.450]  And that's a concern, right?
[31:02.450 --> 31:15.930]  And then also some of the questions, we've already talked about how some of the research questions that are funded are not of interest to indigenous peoples, whereas research questions and questions of health that are of interest are not adequately funded.
[31:15.930 --> 31:21.010]  Like for instance, social programming, which really would be a benefit to indigenous peoples.
[31:21.010 --> 31:24.590]  But then also who's being funded is also of concern.
[31:24.610 --> 31:44.770]  Because oftentimes, and this is my own personal and professional gripe, early career indigenous scientists or even scientists that don't have PhDs, but do community work are often not considered to the level of professional expertise as warranted to be PIs of their own NIH or NSF awards.
[31:45.010 --> 32:12.930]  And unfortunately, under that model, it preferences non-indigenous scientists who perhaps have even gotten bad reputations for merely just recruiting indigenous peoples and studies, but that's purely just for... it starts a cycle of just perpetuating the same sort of extractive unilateral research that benefits professionally and also non-indigenous peoples.
[32:13.070 --> 32:32.290]  And I think just from that standpoint too, I do have to say that when we started the NBDC as an organization, we had a lot of interest from for-profit companies that thought that we just had this huge bank of indigenous peoples' biological information that they wanted access to,
[32:32.290 --> 32:39.670]  but they wanted it particularly to study and create therapeutics that affected the majority population.
[32:39.670 --> 32:47.750]  We pushed back and asked simply, what about using indigenous peoples' data to create therapeutics that affect indigenous peoples?
[32:47.970 --> 32:54.670]  They kind of, first of all, they stopped, well, your people only constitute one to 3% of the population.
[32:54.670 --> 33:01.890]  It's just not commercially viable enough for us to create therapeutics that benefit you.
[33:01.890 --> 33:12.660]  And I'm like, well, well, that shows you that a profit-driven or commercial model or even a capitalistic model to go that far is not going to be...
[33:14.070 --> 33:19.390]  it's not going to derive social benefit for underrepresented peoples either.
[33:20.170 --> 33:20.830]  Yeah.
[33:21.630 --> 33:23.850]  Yeah, that's a great point.
[33:24.010 --> 33:31.070]  So I'd like to offer up a question around, so where do we go from here?
[33:31.070 --> 33:46.850]  How do we change, you know, the current ways of doing things so that there is more respect for persons, respect for data, and, you know, kind of progress in that way towards equity?
[33:48.090 --> 34:14.050]  I kind of want to address one of the elephants in the room, and one is that I share with Larry this altruistic view of science and this possible benefit to people in the planet through research, but that's on our side of the equation, right?
[34:14.050 --> 34:24.650]  There's this invisible membrane between what we do as scientists and what commercial companies do in the public realm, and nobody ever talks about that.
[34:24.650 --> 34:38.450]  Crystal has a wonderful slide that talks about that, where step A is collecting data for the benefit of all mankind, and step C is actually implementing that benefit.
[34:38.450 --> 34:42.810]  But step B, how do you get to that, is always missing.
[34:42.810 --> 34:48.510]  So the implementation often goes to populations who can afford it.
[34:48.510 --> 34:50.910]  Everybody else is left out.
[34:50.910 --> 35:10.990]  So I think one of the steps forward is trying to talk about those kinds of researchers and scientists that do that translation from step A to step C in an ethical, legal, and beneficial manner for those other populations that are not white.
[35:13.530 --> 35:15.910]  Yeah, very well said.
[35:20.400 --> 35:39.280]  Yeah, it's also not just enough to train people, underrepresented scientists, to work in a field that still, you know, reinforces colonial structures of benefit, because we're just...
[35:39.960 --> 36:03.200]  this is actually a huge problem that I faced as a mentee, that mentors were encouraged to train, particularly Native American students, but they had no interest actually in training a student that wanted to push back on how academia has traditionally done things.
[36:03.200 --> 36:16.880]  And we really need to have cognizance of not just training underrepresented community members of scientists, but then also acknowledging that they're advocates for within their own communities.
[36:16.880 --> 36:31.460]  And sometimes we get pushed back as being, since we're representatives of these communities, that we cannot be objective purveyors of the science, and then our viewpoints get undermined in that regards.
[36:31.460 --> 36:36.580]  And then, you know, a non-Indigenous person, for instance, gets privileged.
[36:36.760 --> 36:42.640]  And then it also sets up weird economies of Indigenous scientists being anti-science.
[36:42.640 --> 36:52.700]  But in reality, what we're talking about is more calling to question progress that continues to exploit community members and underrepresenting disenfranchised people.
[36:52.700 --> 37:13.120]  So, I think not only is it important to educate everyone that is utilizing data and creating and producing data and ensuring that they understand what the value is of their data, because even if they don't have value of the data themselves, somebody somewhere out there,
[37:13.120 --> 37:17.020]  I'm sure, is able to commercially co-op that data.
[37:17.020 --> 37:31.780]  But then also, it's also up to researchers and policymakers and, dare I say, legislators to be cognizant of these larger ethical questions related to data as well.
[37:31.780 --> 37:50.320]  Because ultimately, you know, it's where these lines of data ownership versus stewardship and data privacy does require a lot of looking back on how data has been used.
[37:50.320 --> 38:00.680]  And we really have to look to this type of like these harmful histories of extraction in order to ensure that those histories don't get repeated.
[38:02.300 --> 38:07.520]  Yeah, this is going to sound like hyperbole and I'm not picking on you, Larry.
[38:08.220 --> 38:10.800]  Feel free to pick on me, Joe, I can take it.
[38:11.540 --> 38:19.580]  So these are all disruptive technologies, whether it's omics, machine learning, AI, they're all disruptive, right?
[38:19.580 --> 38:35.520]  And the hallmarks of disruptive technologies is that they quickly make other things obsolete, and then they cross these either disciplinary or technological borders, and then they quickly sink into the background, right?
[38:35.520 --> 38:37.300]  So these are really powerful.
[38:37.300 --> 38:44.560]  These are, in my view, world-changing technologies on the same order as nuclear technology, right?
[38:44.560 --> 38:51.740]  So the question is, even though you and I have the greatest intentions, is the world ready for this, the rest of the world?
[38:51.740 --> 38:55.500]  Because certainly people weren't ready for nuclear technology.
[38:55.500 --> 39:06.200]  Even though scientists might have had the best of intentions, the world was not ready for nuclear technology, and we're still having to suffer all the after effects of nuclear technology.
[39:06.200 --> 39:07.900]  And this is the same thing.
[39:07.900 --> 39:12.680]  Can we do this better with these technologies that are going to change the world?
[39:12.680 --> 39:28.120]  As scientists, do we have a bigger obligation, not only because of history of these past science misuses, but also because we are now aware of the power that we are creating as scientists?
[39:29.220 --> 39:31.780]  Yeah, I actually think that's a good question.
[39:31.780 --> 39:34.820]  You led right into the national security argument, Joe.
[39:35.400 --> 39:49.580]  And I, you know, for me, that's one of the things that I think about a lot is that every time you're charting new territory, every time you're developing new technology or new methodologies, you don't know exactly what the potential is there.
[39:49.580 --> 40:02.460]  And I think that, you know, you have to think with security and all of those considerations, you have to think ahead about what that's going to look like, and not wait until, you know, the cat's kind of out of the bag, so to speak.
[40:03.500 --> 40:14.420]  Well, and, you know, I guess I'm going to come down on the side of continued growth in openness and some cultural change.
[40:14.860 --> 40:19.500]  So, you know, Crystal, I agree, I would love legislators to act better.
[40:20.220 --> 40:27.060]  And, you know, decolonizing the world is a big, important task.
[40:27.060 --> 40:28.060]  But I'm a scientist.
[40:28.060 --> 40:35.140]  And so what I can do, what I have the capacity for, is to try to make science better.
[40:35.420 --> 40:42.380]  And I completely agree with you, Crystal, that decolonizing science is going to change the way mentorship works.
[40:42.380 --> 40:50.140]  It's going to change the way that the relationship between Indigenous scientists and the rest of science is going to work.
[40:50.320 --> 40:57.600]  There are some great, important things that people like you and Joe and others are doing to make that happen.
[40:58.200 --> 41:09.600]  And I want to hark back to something Joe said a little earlier about scientists who didn't share their data back in the old days, in the 1900s, back when he and I were your age.
[41:09.600 --> 41:11.620]  That still happens in some cases.
[41:11.620 --> 41:18.860]  Yeah, well, it is the case that, you know, your success in science was about the data you collected, right?
[41:18.860 --> 41:24.380]  I mean, people held their data really tightly, and they'd rather use other people's toothbrushes than other people's data.
[41:24.460 --> 41:26.260]  And those days are gone.
[41:26.260 --> 41:32.760]  And we changed that in large part because of internally generated cultural shifts.
[41:32.840 --> 41:43.580]  Okay, there was an editorial not that long ago in the New England Journal of Medicine, decrying data parasites, people who used other people's data to get new science, right?
[41:43.580 --> 41:44.980]  And we turned that on its head.
[41:44.980 --> 41:47.720]  We made a prize for the best data parasites.
[41:47.720 --> 41:54.440]  And we finally got the New England Journal of Medicine to admit that it was wrong and scheduled this, you know, this whole event around reuse of data.
[41:54.440 --> 41:56.240]  And it's now kind of a thing.
[41:56.240 --> 41:57.580]  So change happens.
[41:57.580 --> 42:06.440]  It is possible to improve our culture, our culture of science, and the culture of the application of science and of technologists.
[42:06.660 --> 42:12.780]  And so I'm hopeful that this, you know, I've been in this openness movement for 30 some odd years.
[42:12.780 --> 42:16.100]  And I started off as a complete evangelical, okay?
[42:16.100 --> 42:17.520]  Everything should be open.
[42:17.520 --> 42:19.640]  There's no reason any of it should be hidden.
[42:19.640 --> 42:22.600]  And I've slowly sort of changed my mind about that.
[42:22.760 --> 42:26.520]  I have been convinced by people like Michelle that there are bad actors in the world.
[42:26.520 --> 42:30.020]  And if everything is open, we will get hurt.
[42:30.020 --> 42:45.580]  And I've been convinced by people like you, Joe, that openness in the past has often been exploitive and colonial and has not benefited the people whose good faith and contribution of their selves and their data did not end up benefiting them.
[42:45.580 --> 42:48.680]  So there's definitely things we can do better.
[42:48.840 --> 43:05.760]  But cultural change within science, I think, has been a really wonderful way of, during my lifetime, of making science better, more reproducible, more answerable to the needs of broader communities, and more open to more people to do it.
[43:05.760 --> 43:13.420]  And so I think that where we need to go is to continue to have these discussions and to push for cultural change, both within and beyond science.
[43:14.740 --> 43:28.160]  So I'm really happy about our organization, because we're really encouraging communities to take ownership of their data and recognizing that they have the capacity to do so.
[43:28.160 --> 43:34.680]  But that also requires a huge... requires a lot.
[43:34.680 --> 43:54.660]  We just fundamentally, at this point, don't have enough trained individuals who are local to our rural tribal communities to undertake and take the data science tools and the data and to build data STEM economies.
[43:54.660 --> 44:19.980]  And that's what Joe and I are really trying to push for, is rather than having this one way move of data exiting our communities, and then waiting, praying, hoping that one day it will be returned to us in some beneficial manner, that we actually continually think about training the next leaders from our communities to keep the data held locally.
[44:20.080 --> 44:22.860]  Now, this is great.
[44:22.860 --> 44:38.600]  And I see we're nodding our heads, but some people who are not interested in these types of movements are now stating that this is perhaps, again, anti to progress, anti to the openness of science.
[44:38.700 --> 44:54.540]  And I question whether or not their motive is, again, for the benefit of society and for improving equity in science, and versus they're just seeing it as a hindrance to their own access to the data.
[44:54.600 --> 45:24.740]  So I am personally, you know, cautiously trying to build awareness that, you know, there are actors, scientific actors that are really self-interested, self-motivated, and that we need to ensure that we are also at the same time building up a next generation of data leaders that understand that there's these challenges in place,
[45:24.740 --> 45:28.680]  not only technologically, but also within their own fields.
[45:28.680 --> 45:37.180]  And I'm really hoping that we have a cognizance of mentors to allow us to be able to do so.
[45:37.180 --> 45:49.580]  Otherwise, you know, it would take very little to squash the academic trajectories of Indigenous scientists like myself and others.
[45:51.420 --> 46:15.300]  I wanted to go back again to the excitement that Larry has about changing culture, and I agree that changing the view of scientists and researchers is important, but again, we do not have that much clout, and that was demonstrated during the COVID pandemic.
[46:15.300 --> 46:19.460]  Scientists were not listened to because we're not the decision makers.
[46:19.500 --> 46:21.680]  We are not the decision makers.
[46:22.080 --> 46:32.200]  And open data and open access, in my experience since about 2014, has violated the common rule.
[46:32.900 --> 46:41.260]  The research should not be coercive, but they identify that coerciveness individually, not as a group.
[46:41.520 --> 46:50.960]  And because of the open data ideal, you now have private funders insisting that you make your data open.
[46:50.960 --> 46:54.880]  You have journals that insist on making your data open.
[46:55.100 --> 46:59.100]  And of course, the public funders want to make your data open.
[46:59.100 --> 47:00.340]  And this is coercive.
[47:00.340 --> 47:08.820]  So tribes are sort of left in this lurch where they need the research, and they don't want to share their data.
[47:08.820 --> 47:15.060]  So they either have to capitulate to one or the other to their detriment, right?
[47:15.060 --> 47:19.000]  If they don't get the research, they're not going to have implementation.
[47:19.120 --> 47:24.100]  If they do do the research, they're probably not going to have implementation.
[47:24.140 --> 47:33.080]  And then it allows other people to capitalize on whatever unique data that they produce, because they can't capitalize on it for themselves, right?
[47:33.080 --> 47:44.600]  And so I think that this question about open data, as Crystal said earlier, has to really be interrogated contextually on a case-by-case basis.
[47:45.360 --> 47:49.380]  Yeah, I think that's a great point to end on.
[47:49.380 --> 47:52.200]  You know, I think this has been a wonderful discussion.
[47:52.200 --> 47:56.000]  We're definitely not solving any problems with this discussion.
[47:56.000 --> 47:59.240]  But at the same time, I think it's really important to continue the conversation.
[47:59.240 --> 48:01.680]  So thank you, everybody, for listening.
[48:01.680 --> 48:04.940]  Thank you so much, Crystal, Larry, and Joe for participating.
[48:05.240 --> 48:07.560]  And that's a wrap.