6 ms·

The Overlooked Potential of African DNA

39 points by sethbannon 7y ago

JoeAltmaier 7y ago
All sounded good until the part about only collecting samples in Nigeria. All the world has less genetic variation, than that between any two populations in Africa. Sampling one African population is a start, in that it nearly doubles the DNA sample space for study. But it won't help all of Africa - just the one population sampled.
- SideburnsOfDoom 7y ago
  > Sampling one African population is a start, Nigeria has numerous ethnic groups present in it; but yes, I hope they range further afield as well.
- bayesian_horse 7y ago
  According to the article, Nigeria has about 500 distinct ethnic groups. And I guess the biggest problem is not so much getting samples for the DNA, but rather to link it to phenotype data. And that will be highly country-specific.
- JadeNB 7y ago
  > All the world has less genetic variation, than that between any two populations in Africa. I guess this is just a flight of poesy, but it seems like a literally impossible statement.
  - SideburnsOfDoom 7y ago
    I think it's possible if it's actually "All _the rest of_ the world has less genetic variation, than that between any two populations in Africa." I don't know if that's true, but it's true that Africans have more genetic variation than anyone else on Earth. http://www.nbcnews.com/id/30502963/ns/technology_and_science-science/t/africans-have-worlds-greatest-genetic-variation/ http://www.nbcnews.com/id/30502963/ns/technology_and_science... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4067985/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4067985/ > most genetic variation in non-African populations is a subset of the variation found in African populations https://www.ashg.org/education/pdf/geneticvariation.pdf https://www.ashg.org/education/pdf/geneticvariation.pdf > Larger Genetic Differences Within Africans Than Between Africans and Eurasians https://www.genetics.org/content/161/1/269 https://www.genetics.org/content/161/1/269
  - thrower123 7y ago
    We can thank the ubiquitous Mercator projection for us having no idea how big Africa actually is. https://blogs.scientificamerican.com/observations/africa-is-way-bigger-than-you-think/ https://blogs.scientificamerican.com/observations/africa-is-...
    - carapace 7y ago
      World map of human migrations on a Dymaxion projection (no visible distortion of landmass size/shape.) https://en.wikipedia.org/wiki/File:World_map_of_prehistoric_human_migrations.jpg https://en.wikipedia.org/wiki/File:World_map_of_prehistoric_...
    - JadeNB 7y ago
      My issue wasn't with the size of Africa (though I have seen that visualisation before, and find it startling as intended), but with the fact that Africa, huge as it is, is part of the world, so that its genetic diversity is part of the world's.
  - jjtheblunt 7y ago
    Yes, its wording is not quite right.
    - jjtheblunt 7y ago
      In particular, the assertion I knew from working in human genomics was that, if one were to choose two distinct populations A and B, with the greatest diversity between the populations available on the planet, then both A and B would occur in Africa.
- sethbannon 7y ago
  54Gene investor here. While the team is starting in Nigeria, they absolutely plan to expand to the whole of Africa. The logistics, regulations, and cultural norms differ country by country in Africa, so it's easiest to start in a single country (Nigeria) and expand from there.
  - Zod666 7y ago
    I find this very interesting. I wanted to ask how you invested and if that was still possible for others? Was it private investing?
- jjtheblunt 7y ago
  "than that between any two populations in Africa" : i think you've worded that incorrectly, and please correct me if i misunderstand. I thought the correct assertion was that there is greater diversity in Africa itself than occurs in the world outside Africa. What you wrote could be fooled by choosing two very proximal populations within Africa, thereby failing to take advantage of that diversity.
equalunique 7y ago
In the US, would genetic information like the kind collected here be covered by HIPAA? Are there any similar laws/regulations/protections in place for Africans? Is there something unique about those protections (or lack therof) when it comes to Nigeria, where 54gene is operating? I'm skeptical.
- droithomme 7y ago
  I've had pretty heated arguments over this with an old friend of mine from one of the big US companies. They argue forcefully that DNA information is not medical data and HIPAA does not apply to them at all. I don't agree with their claim, and believe HIPAA should apply, but it seems the US regulators are going along with them for now.
  - Nasrudith 7y ago
    They may be technically correct for their company and applications. Height and weight are medical data but a leak of that from non-medical contexts wouldn't be a HIPAA breach. Your doctor exposing it in a data breach would be in deep shit.
  - carapace 7y ago
    Henrietta Lacks at scale. https://en.wikipedia.org/wiki/HeLa https://en.wikipedia.org/wiki/HeLa https://en.wikipedia.org/wiki/Henrietta_Lacks https://en.wikipedia.org/wiki/Henrietta_Lacks
daemonk 7y ago
Are they selling access to the DNA biobank or are they generating sequencing data and selling access to the data? If you guys are generating sequencing data, how are you guys doing it? I would assume you guys are actually sequencing the DNA and not using a microarray? It would seem kind of pointless to use a microarray designed for the current eurocentric data on these samples. What coverage are you sequencing these samples at?

39 points by sethbannon 7y ago

All sounded good until the part about only collecting samples in Nigeria. All the world has less genetic variation, than that between any two populations in Africa. Sampling one African population is a start, in that it nearly doubles the DNA sample space for study. But it won't help all of Africa - just the one population sampled.

SideburnsOfDoom 7y ago
> Sampling one African population is a start, Nigeria has numerous ethnic groups present in it; but yes, I hope they range further afield as well.
bayesian_horse 7y ago
According to the article, Nigeria has about 500 distinct ethnic groups. And I guess the biggest problem is not so much getting samples for the DNA, but rather to link it to phenotype data. And that will be highly country-specific.
JadeNB 7y ago
> All the world has less genetic variation, than that between any two populations in Africa. I guess this is just a flight of poesy, but it seems like a literally impossible statement.
- SideburnsOfDoom 7y ago
  I think it's possible if it's actually "All _the rest of_ the world has less genetic variation, than that between any two populations in Africa." I don't know if that's true, but it's true that Africans have more genetic variation than anyone else on Earth. http://www.nbcnews.com/id/30502963/ns/technology_and_science-science/t/africans-have-worlds-greatest-genetic-variation/ http://www.nbcnews.com/id/30502963/ns/technology_and_science... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4067985/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4067985/ > most genetic variation in non-African populations is a subset of the variation found in African populations https://www.ashg.org/education/pdf/geneticvariation.pdf https://www.ashg.org/education/pdf/geneticvariation.pdf > Larger Genetic Differences Within Africans Than Between Africans and Eurasians https://www.genetics.org/content/161/1/269 https://www.genetics.org/content/161/1/269
- thrower123 7y ago
  We can thank the ubiquitous Mercator projection for us having no idea how big Africa actually is. https://blogs.scientificamerican.com/observations/africa-is-way-bigger-than-you-think/ https://blogs.scientificamerican.com/observations/africa-is-...
  - carapace 7y ago
    World map of human migrations on a Dymaxion projection (no visible distortion of landmass size/shape.) https://en.wikipedia.org/wiki/File:World_map_of_prehistoric_human_migrations.jpg https://en.wikipedia.org/wiki/File:World_map_of_prehistoric_...
  - JadeNB 7y ago
    My issue wasn't with the size of Africa (though I have seen that visualisation before, and find it startling as intended), but with the fact that Africa, huge as it is, is part of the world, so that its genetic diversity is part of the world's.
- jjtheblunt 7y ago
  Yes, its wording is not quite right.
  - jjtheblunt 7y ago
    In particular, the assertion I knew from working in human genomics was that, if one were to choose two distinct populations A and B, with the greatest diversity between the populations available on the planet, then both A and B would occur in Africa.
sethbannon 7y ago
54Gene investor here. While the team is starting in Nigeria, they absolutely plan to expand to the whole of Africa. The logistics, regulations, and cultural norms differ country by country in Africa, so it's easiest to start in a single country (Nigeria) and expand from there.
- Zod666 7y ago
  I find this very interesting. I wanted to ask how you invested and if that was still possible for others? Was it private investing?
jjtheblunt 7y ago
"than that between any two populations in Africa" : i think you've worded that incorrectly, and please correct me if i misunderstand. I thought the correct assertion was that there is greater diversity in Africa itself than occurs in the world outside Africa. What you wrote could be fooled by choosing two very proximal populations within Africa, thereby failing to take advantage of that diversity.

equalunique 7y ago

In the US, would genetic information like the kind collected here be covered by HIPAA? Are there any similar laws/regulations/protections in place for Africans? Is there something unique about those protections (or lack therof) when it comes to Nigeria, where 54gene is operating? I'm skeptical.

droithomme 7y ago
I've had pretty heated arguments over this with an old friend of mine from one of the big US companies. They argue forcefully that DNA information is not medical data and HIPAA does not apply to them at all. I don't agree with their claim, and believe HIPAA should apply, but it seems the US regulators are going along with them for now.
- Nasrudith 7y ago
  They may be technically correct for their company and applications. Height and weight are medical data but a leak of that from non-medical contexts wouldn't be a HIPAA breach. Your doctor exposing it in a data breach would be in deep shit.
- carapace 7y ago
  Henrietta Lacks at scale. https://en.wikipedia.org/wiki/HeLa https://en.wikipedia.org/wiki/HeLa https://en.wikipedia.org/wiki/Henrietta_Lacks https://en.wikipedia.org/wiki/Henrietta_Lacks

daemonk 7y ago

Are they selling access to the DNA biobank or are they generating sequencing data and selling access to the data? If you guys are generating sequencing data, how are you guys doing it? I would assume you guys are actually sequencing the DNA and not using a microarray? It would seem kind of pointless to use a microarray designed for the current eurocentric data on these samples. What coverage are you sequencing these samples at?