AGENDA

REWS Diversity in Datasets Meeting 3
28th July 2022 @ 2pm UTC

(dial in numbers below)

Zoom Recording [password: 66YM7*N!]

Zoom Details:

Join Zoom Meeting

https://us02web.zoom.us/j/81844677427?pwd=WFE3UGNFY01ZUnpMeVBubldIUmtwQT09

Meeting ID: 818 4467 7427

Passcode: 492987

Attendees:

1. Maili Raven-Adams
2. Lindsay Smith
3. Ariana Gonzalez
4. Hanshi Liu
5. Gabriela Chavarria
6. Anna Lewis
7. Elena Ghanaim
8. Sue Chen
9. Diksha Srivastava
10. Anne de Hond
11. Beatrice Kaiser
12. Alham Saadat
13. Norbert Tavares
14. Bimal Chaudhari
15. Shuba Gopal
16. Tina Hernandez-Boussard

Minutes:

AL: Last month, we talked about what makes genomics special. Today, we want to focus on the international dimensions, then on the next call, we want to think about what some policy points might be.

Summarizing last month’s points: What makes genomics special?

1. The existence of genomic exceptionalism, and the empirical fact it exists whether or not we think it's justified.
2. Genetics and Genomics has a potential to put a scientific definition of racism. Genetic ancestry blurs into race.
3. The belief that what we found out from genetics is universal about the human species.
4. In genomics, we do a lot of data sharing, including across international borders.  

AS: The idea of commodification of genetic and genomics data that large companies have, that they are selling to other organizations. A bit different from other healthcare data. The way that it is sold and collected, there’s not always clarity on other uses of it. Other types of health data are more protected.

GC: Agree that the widespread international data sharing is something very particular. You can be identified, which is not necessarily the case with other data.

AL: In your own jurisdiction - are there commonalities or differences about genetic data being identifying information?

AG: Gray space in Argentina where that data is from no one and everyone. Data is often sent abroad for sequencing/analysis. That data doesn’t come back. Lack of responsibility in our region to keep safe or create regulatory laws about it. No entity to tell us whether we are doing something right.

HL: We’re already making a bit of an assumption, which is that it’s always health data. Should be careful about distinguishing that.

AL: Yes, like forensic applications.

SC: In the US, we do have laws and regulations that protect health data. There is a huge compliance piece that goes along with that. Part of our granting process includes applicants having to develop a data sharing plan. Reviewing de-identification. NIH is also very concerned with regards to privacy not just re-identification but there are also considerations in place regarding individual and group harms.

NT: It varies widely in the US. If the data are coming from hospitals, they’re very conservative and there are concerns about re-identification. Can’t get individual level data, it’s always going to be in aggregate. Single cell biology is a little bit different, in that we often don’t want the genetics data but it’s generated. Can’t identify anyone today, but will be able to later.

AL: Genetics data can potentially be everywhere.

BC: Genomic data in healthcare is not the biggest privacy threat or most identifiable thing. It takes an enormous effort to identify a single human using genetic data.

AL: The privacy is not primary, I agree.

HL: The idea of identifiability and health data. There is still a lot we don’t know about genetics. The information we think as “junk” data, may be identifiable later.

International Dimensions

MRA: Take the broad idea and situate in international scope. Recognize that all of these international institutes are doing things differently. One of the first things we were going to ask was the international elements of exceptionalism, which we’ve touched on already. Anything else to add?

NT: In my experience, some countries are very restrictive on the ability to export data out.

AL: To give an example from the US, US researchers are trying to start projects based in Africa, to help African-Americans in the US. Idea that we need data from over there.

BC: If you’re looking for non-genetic example, international HIV trials that were done in Africa/India and sold to Europeans.

MRA: Links to another issue about benefit sharing and recognition.

THB: There is a lot of effort regarding clinical data to try and capture more the electronic data in other places to address fair data sharing and data diversity. The way we think of diversity is different from how other countries might think of it.

AL: Is that in order to benefit people in those countries?

THB: Good question, its definitely led by US and other international researchers. We have a lot of programs now that are based in other countries to exchange students and professors to not just use data but build expertise.

GC: There has been distrust towards international collaboration in Costa Rica because you lose all control of your data. We get many offers to collaborate, and have to be careful to choose collaborations that are equal. Genotyping is always done abroad. Need to make sure we have access to our raw data back.

AL: You’re seeing more collaborations because of diversity, when do they approach you? How do they spell it out?

GC: Usually it’s diversity, as a latin american country. We do have a long history of genomic research so we have the infrastructure. We have a national health system where theoretically you can access health information.

NT: There are a couple of issues. It’s not just diversity in certain locations, it’s the infrastructure. Many countries are not approached for that reason. That lack of capacity is a main barrier. 10x has one service agent for the entire continent of South America.

AS: Reiterating the importance of doing international collection of data and samples and need to have bi-directional resource building.

THB: Think about policies to be established around that.

HL: Helicopter science or helicopter research. Go to a place and do their research and leave, without leaving any resource or capacity building.

AL: Can we think about where the drive for diversity comes from. The people who are calling for diversity, why?

MRA: I think about applicable research, need the research applicable and inclusive to those being studied.

THB: Journals and funders now require you have diversity plans and statements.

GC: Agree with Tina, it’s politics and what researchers are forced to do. Good opportunity for us.

AL: There is often an ambiguity whether “genomics for everyone” means just my country, or really everywhere.

BC [in chat]: There is a smaller group of clinicians for whom the drive is the simple fact that African genomes are necessary to clarify allele frequencies and what is rare

NT: It’s just good science. Push has been recent papers that shows that majority of data comes from Europeans. The way funding works in the NIH and the UK, have to connect it back to something for the US and the UK.

BC: It’s also good medicine. I’m also a consumer of this data, and the cynical truth is that one extra African genome is worth 10k european genomics. I want as many African genomes as possible to clarify allele frequencies.  

AL: What does diversity mean, should revisit this as it’s just not representation.

Next Steps:

MRA: As a bit of homework, how to address these different tensions and think about the policy elements we want to cover and how to address these.

THB: For our next meeting, all think about the policies we could ensure that you feel you’re getting equal benefit.

AL: GA4GH has many policies about data sharing, so we really need to focus on the diversity piece

AHB: There’s a difference in helicopter science

MRA: Definitely something to consider because it’s a key issue

AL: Also start next meeting with getting on the same page about what is diversity.

MRA: I’ll circulate the document we started with the key ideas, feel free to add thoughts. Hoping to schedule the next meeting late August/early September.