5 ms·

The major genetic risk factor for severe Covid-19 is inherited from Neandertals

51 points by drocer88 6y ago

zrail 6y ago
Standard disclaimer that this is a preprint that has not been peer reviewed.
drocer88 6y ago
Svante Pääbo , one of the preprint authors, is the head of the Max Planck Institute for Evolutionary Anthropology. https://en.wikipedia.org/wiki/Svante_P%C3%A4%C3%A4bo https://en.wikipedia.org/wiki/Svante_P%C3%A4%C3%A4bo . He is known for his work in assembling the Neanderthal and Denisovan genomes. Of particular note in this paper, the genetic variants in the genomic locus discussed are found in 30% of South Asian populations: India and Bangladesh. The highest frequency occurs in Bangladesh, where more than half the population (63%) carries at least one copy of the Neandertal risk variant and 13% is homozygous for the variant. The risk variant is 8% frequency in Europe. The hg19 location of Genetic variants in LD (r2>0.99) with rs11385942 in Eurasian populations can be viewed here at the UCSC genome browser: https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg19&position=chr3%3A45871908%2D45909024 https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg19&position=ch... The genome wide association study that points out this area of the genome is here : https://www.nejm.org/doi/full/10.1056/NEJMoa2020283 https://www.nejm.org/doi/full/10.1056/NEJMoa2020283 . At this "locus 3p21.31, the association signal spanned the genes SLC6A20, LZTFL1, CCR9, FYCO1, CXCR6 and XCR1." Three of the genes, XCR1 CXCR6 CCR9, are involved with chemokine receptor activity and G-protein coupled chemoattractant receptor activity.
- mirimir 6y ago
  Also only ~4% in the US, and virtually absent in Africa.
rossdavidh 6y ago
"Here, we show that the risk is conferred by a genomic segment of ~50 kb that is inherited from Neandertals and occurs at a frequency of ~30% in south Asia and ~8% in Europe." Hmmm...thus far, the mortality per capita from Covid-19 has been much higher in Europe, especially western Europe, than in south Asia. Not saying that couldn't change, but the weekly mortality peaks in western Europe were literally an order of magnitude higher than for Bangladesh, India, or Pakistan. So either this finding is flawed in some way, or (perhaps more likely) even the biggest genetic risk factor for severe Covid-19 is not that big of a factor, and what's going on is mostly not genetic.
- slicktux 6y ago
  Or maybe the stats used for the data in European countries are flawed...or there is less testing in the latter countries....
  - roywiggins 6y ago
    You don't need testing, just accurate body counts. You can see COVID hit Europe as a spike in excess deaths, even if you weren't testing. If there were currently a mortality spike in Bangladesh even higher than Europe's, it would not go unnoticed. Disappearing bodies is not that easy.
    - ethanwillis 6y ago
      I mean.. China's official numbers are around 84k infected.. and much less dead than that. So...
      - ashtonkem 6y ago
        The “count the bodies” technique is meant for situations where the country can’t test for COVID, either because it was too early or they don’t have the infrastructure. That technique won’t work on countries just willing to lie about it.
herf 6y ago
If you have 23andme, the exact allele is not on the chip. However, a different allele is strongly associated with the region: https://you.23andme.com/tools/data/?query=rs10490770&filter_by_platforms=true https://you.23andme.com/tools/data/?query=rs10490770&filter_... The major A/A variant is listed as T/T on 23andme. If you have a result other than "T/T", you're likely to be in the high-risk group. refs: https://www.snpedia.com/index.php/Rs11385942 https://www.snpedia.com/index.php/Rs11385942 [not on 23andme] https://www.snpedia.com/index.php/Rs10490770 https://www.snpedia.com/index.php/Rs10490770
- modeless 6y ago
  That 23andme link doesn't work for me, it says no genes or markers found for rs10490770.
  - drocer88 6y ago
    Try this : https://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?do_not_redirect&rs=rs10490770 https://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?do_not... This breaks down the population frequencies. Of note: the HAPMAP AA frequencies: Gujarati Indians in Houston Texas : 38% , African ancestry in Southwest USA: 4%. It is absent in Chinese, Japanese and sub-saharan Africans.
  - helpme23 6y ago
    I couldn’t get it to work either, did anyone figure out how to view it on there?
- LegitShady 6y ago
  This is actually the first time the 23andme results did anything useful for me. Thank!
  - helpme23 6y ago
    How did you get it to pull on 23 and me. I have more Neanderthal variants than 99% of users so I’m super interested. Thank you for any help you can provide.
not2b 6y ago
But how can this be, given that African-Americans, with far less Neanderthal ancestry than European-Americans, are dying of COVID-19 at a much higher rate in the US?
- xikrib 6y ago
  Given that there's a strong correlation between race and poverty in the US (via racism) and a strong correlation between poverty and infection rate which, it seems to mean poverty is a greater indicator than genetics. After all the genetic factor only comes into play after exposure, while poverty increases the likelyhood of exposure
  - drocer88 6y ago
    Other factors may be obesity ( https://www.minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvlid=25 https://www.minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvl... ) , diabetes ( https://www.minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvlid=18 https://www.minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvl...) and Vitamin D deficiency ( http://www.cooperinstitute.org/2019/09/24/african-americans-at-greatest-risk-of-vitamin-d-deficiency http://www.cooperinstitute.org/2019/09/24/african-americans-... ).
    - IncRnd 6y ago
      What underlies all of that is the strength of the immune system, which will be low for many diets and lifestyles.
  - jjtheblunt 6y ago
    is NYC with its highest rate of cases in the USA a generalization: density of people (which happens in poverty and in opulence, for different reasons) increases likelihood of exposure?
    - TMWNN 6y ago
      Population density really doesn't matter that much for a pandemic as infectious as COVID19. The number of people who live in the mountains or on a farm and only come into town every six months is very, very small. The two big causes of massive early numbers of deaths, including in the northeast US (especially NYC, and NY state, NJ, and Philadelphia to a slightly lesser degree) and places like northern Italy, were a) sending sick old people into old age homes. Other cities/states by and large did not do this. This drastically raised the average age of those who were infected in NYC, and caused the far higher death rates. 42% (!) of US deaths occurred in old age homes. Old age homes are pretty similar everywhere in terms of layout and design, regardless of local population density. b) Assuming that ventilators should be used on every serious case, causing both damage from overoxygenation to healthy lung sacs and overdependence on mechanical respiration that doctors are having a tough time figuring out how to wean patients from. Those that warned three months ago (https://www.statnews.com/2020/04/08/doctors-say-ventilators-overused-for-covid-19/ https://www.statnews.com/2020/04/08/doctors-say-ventilators-...) against ventilator overuse were completely vindicated in retrospect. This happened nationwide but most proportionately in NYC (in part because it was hit hard early so doctors were pioneers). The big Western European countries hit hardest by the pandemic have 40-60 deaths per 10,000; the US is in the 30s, while states like Texas are in the single digits. NYC's death rate is 161 (!!); it singlehandedly prevented the US's death toll from being almost as low as Germany's. NYC's super-high death rate isn't replicated in the same way in other densely populated states that did not make the mistake of accidentally encouraging spreading of the disease among the elderly. Chicago, DC, and Boston were not hit anywhere as hard as NYC was. Look, for example, at the NY Times dataset (https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html https://www.nytimes.com/interactive/2020/us/coronavirus-us-c...). NYC's daily deaths peaked at about 10% of the cases, so we'll say 10% case fatality rate. Chicago's CFR is about 4%. Philadelphia's CFR is close to NYC, but Boston's is 6% and DC's is also about 4%. Basically, the NYC-Philly corridor handled this worse than anywhere else in the country. Again, old age homes contributed 42% of COVID19 deaths, and largely due to big mistakes made in a portion of the northeast US. Those mistakes are no longer being made anywhere, nor is the "putting everyone on ventilators" mistake. This (plus the massive expansion in testing causing the average age of those testing positive to plummet) is why, although southern and western states' cases are rising (or in Los Angeles's case, never stopped rising from the start), deaths are not, and why national daily deaths continue to fall straight down from the mid-April peak.
gorkish 6y ago
I do not have any particular comment on this gene or the topic of COVID-19, but I have a related question for the experts out there. Let's say I were to send off to Dante labs and get my entire genome sequenced. Given emerging research like this where someone says "we have identified this specific gene that does so-and-so," does the average geek have any hope of being able to determine if they have a copy of that gene? Or does that type of interpretation require the effort of a trained expert? How close are we to receiving a personalized genetic vulnerability feed? I'm sure the hypochondriacs out there would sign up in a heartbeat.
tbinmd 6y ago
So in my 23andme it shows me as having the T/T genotype of this marker. Does that mean I do or do not have the at risk marker?
- boberto123 6y ago
  I think you're good. From what I've read, the high risk is C. T/T is modern human. C is neanderthal. So C/C - youre 3 times more likely to get sick. C/T youre 1.5times more likely to get sick. Reference: https://www.gnxp.com/WordPress/2020/07/04/using-23andme-ancestry-family-tree-dna-to-identify-risk-allele-for-respiratory-failure-with-covid-19/comment-page-1/ https://www.gnxp.com/WordPress/2020/07/04/using-23andme-ance... === If this is wrong, someone correct me, but T/T results from 23andme seem to be 84.5% of the population (caucasion people), so YOU SHOULD BE FINE.
boberto123 6y ago
[Sars-Cov-2][Outcome risk] rs10490770 (Caucasisn only) Risk: AG, GG (~15.5% of population?) Safe: AA (~84.5% of population?) --- SO, if 23andme says its T/T and its actually A/A(as herf said), than 84.5% of the population would have T/T via 23andme. I am confused then. If I have T/T through 23andme, am I at higher risk when ~85% of the population has the same thing?
nigit 6y ago
Covid's no big deal cause I already got aids

51 points by drocer88 6y ago

Standard disclaimer that this is a preprint that has not been peer reviewed.

Svante Pääbo , one of the preprint authors, is the head of the Max Planck Institute for Evolutionary Anthropology. https://en.wikipedia.org/wiki/Svante_P%C3%A4%C3%A4bo https://en.wikipedia.org/wiki/Svante_P%C3%A4%C3%A4bo . He is known for his work in assembling the Neanderthal and Denisovan genomes. Of particular note in this paper, the genetic variants in the genomic locus discussed are found in 30% of South Asian populations: India and Bangladesh. The highest frequency occurs in Bangladesh, where more than half the population (63%) carries at least one copy of the Neandertal risk variant and 13% is homozygous for the variant. The risk variant is 8% frequency in Europe. The hg19 location of Genetic variants in LD (r2>0.99) with rs11385942 in Eurasian populations can be viewed here at the UCSC genome browser: https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg19&position=chr3%3A45871908%2D45909024 https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg19&position=ch... The genome wide association study that points out this area of the genome is here : https://www.nejm.org/doi/full/10.1056/NEJMoa2020283 https://www.nejm.org/doi/full/10.1056/NEJMoa2020283 . At this "locus 3p21.31, the association signal spanned the genes SLC6A20, LZTFL1, CCR9, FYCO1, CXCR6 and XCR1." Three of the genes, XCR1 CXCR6 CCR9, are involved with chemokine receptor activity and G-protein coupled chemoattractant receptor activity.

mirimir 6y ago
Also only ~4% in the US, and virtually absent in Africa.

rossdavidh 6y ago

"Here, we show that the risk is conferred by a genomic segment of ~50 kb that is inherited from Neandertals and occurs at a frequency of ~30% in south Asia and ~8% in Europe." Hmmm...thus far, the mortality per capita from Covid-19 has been much higher in Europe, especially western Europe, than in south Asia. Not saying that couldn't change, but the weekly mortality peaks in western Europe were literally an order of magnitude higher than for Bangladesh, India, or Pakistan. So either this finding is flawed in some way, or (perhaps more likely) even the biggest genetic risk factor for severe Covid-19 is not that big of a factor, and what's going on is mostly not genetic.

slicktux 6y ago
Or maybe the stats used for the data in European countries are flawed...or there is less testing in the latter countries....
- roywiggins 6y ago
  You don't need testing, just accurate body counts. You can see COVID hit Europe as a spike in excess deaths, even if you weren't testing. If there were currently a mortality spike in Bangladesh even higher than Europe's, it would not go unnoticed. Disappearing bodies is not that easy.
  - ethanwillis 6y ago
    I mean.. China's official numbers are around 84k infected.. and much less dead than that. So...
    - ashtonkem 6y ago
      The “count the bodies” technique is meant for situations where the country can’t test for COVID, either because it was too early or they don’t have the infrastructure. That technique won’t work on countries just willing to lie about it.

herf 6y ago

If you have 23andme, the exact allele is not on the chip. However, a different allele is strongly associated with the region: https://you.23andme.com/tools/data/?query=rs10490770&filter_by_platforms=true https://you.23andme.com/tools/data/?query=rs10490770&filter_... The major A/A variant is listed as T/T on 23andme. If you have a result other than "T/T", you're likely to be in the high-risk group. refs: https://www.snpedia.com/index.php/Rs11385942 https://www.snpedia.com/index.php/Rs11385942 [not on 23andme] https://www.snpedia.com/index.php/Rs10490770 https://www.snpedia.com/index.php/Rs10490770

modeless 6y ago
That 23andme link doesn't work for me, it says no genes or markers found for rs10490770.
- drocer88 6y ago
  Try this : https://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?do_not_redirect&rs=rs10490770 https://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?do_not... This breaks down the population frequencies. Of note: the HAPMAP AA frequencies: Gujarati Indians in Houston Texas : 38% , African ancestry in Southwest USA: 4%. It is absent in Chinese, Japanese and sub-saharan Africans.
- helpme23 6y ago
  I couldn’t get it to work either, did anyone figure out how to view it on there?
LegitShady 6y ago
This is actually the first time the 23andme results did anything useful for me. Thank!
- helpme23 6y ago
  How did you get it to pull on 23 and me. I have more Neanderthal variants than 99% of users so I’m super interested. Thank you for any help you can provide.

not2b 6y ago

But how can this be, given that African-Americans, with far less Neanderthal ancestry than European-Americans, are dying of COVID-19 at a much higher rate in the US?

xikrib 6y ago
Given that there's a strong correlation between race and poverty in the US (via racism) and a strong correlation between poverty and infection rate which, it seems to mean poverty is a greater indicator than genetics. After all the genetic factor only comes into play after exposure, while poverty increases the likelyhood of exposure
- drocer88 6y ago
  Other factors may be obesity ( https://www.minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvlid=25 https://www.minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvl... ) , diabetes ( https://www.minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvlid=18 https://www.minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvl...) and Vitamin D deficiency ( http://www.cooperinstitute.org/2019/09/24/african-americans-at-greatest-risk-of-vitamin-d-deficiency http://www.cooperinstitute.org/2019/09/24/african-americans-... ).
  - IncRnd 6y ago
    What underlies all of that is the strength of the immune system, which will be low for many diets and lifestyles.
- jjtheblunt 6y ago
  is NYC with its highest rate of cases in the USA a generalization: density of people (which happens in poverty and in opulence, for different reasons) increases likelihood of exposure?
  - TMWNN 6y ago
    Population density really doesn't matter that much for a pandemic as infectious as COVID19. The number of people who live in the mountains or on a farm and only come into town every six months is very, very small. The two big causes of massive early numbers of deaths, including in the northeast US (especially NYC, and NY state, NJ, and Philadelphia to a slightly lesser degree) and places like northern Italy, were a) sending sick old people into old age homes. Other cities/states by and large did not do this. This drastically raised the average age of those who were infected in NYC, and caused the far higher death rates. 42% (!) of US deaths occurred in old age homes. Old age homes are pretty similar everywhere in terms of layout and design, regardless of local population density. b) Assuming that ventilators should be used on every serious case, causing both damage from overoxygenation to healthy lung sacs and overdependence on mechanical respiration that doctors are having a tough time figuring out how to wean patients from. Those that warned three months ago (https://www.statnews.com/2020/04/08/doctors-say-ventilators-overused-for-covid-19/ https://www.statnews.com/2020/04/08/doctors-say-ventilators-...) against ventilator overuse were completely vindicated in retrospect. This happened nationwide but most proportionately in NYC (in part because it was hit hard early so doctors were pioneers). The big Western European countries hit hardest by the pandemic have 40-60 deaths per 10,000; the US is in the 30s, while states like Texas are in the single digits. NYC's death rate is 161 (!!); it singlehandedly prevented the US's death toll from being almost as low as Germany's. NYC's super-high death rate isn't replicated in the same way in other densely populated states that did not make the mistake of accidentally encouraging spreading of the disease among the elderly. Chicago, DC, and Boston were not hit anywhere as hard as NYC was. Look, for example, at the NY Times dataset (https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html https://www.nytimes.com/interactive/2020/us/coronavirus-us-c...). NYC's daily deaths peaked at about 10% of the cases, so we'll say 10% case fatality rate. Chicago's CFR is about 4%. Philadelphia's CFR is close to NYC, but Boston's is 6% and DC's is also about 4%. Basically, the NYC-Philly corridor handled this worse than anywhere else in the country. Again, old age homes contributed 42% of COVID19 deaths, and largely due to big mistakes made in a portion of the northeast US. Those mistakes are no longer being made anywhere, nor is the "putting everyone on ventilators" mistake. This (plus the massive expansion in testing causing the average age of those testing positive to plummet) is why, although southern and western states' cases are rising (or in Los Angeles's case, never stopped rising from the start), deaths are not, and why national daily deaths continue to fall straight down from the mid-April peak.

gorkish 6y ago

I do not have any particular comment on this gene or the topic of COVID-19, but I have a related question for the experts out there. Let's say I were to send off to Dante labs and get my entire genome sequenced. Given emerging research like this where someone says "we have identified this specific gene that does so-and-so," does the average geek have any hope of being able to determine if they have a copy of that gene? Or does that type of interpretation require the effort of a trained expert? How close are we to receiving a personalized genetic vulnerability feed? I'm sure the hypochondriacs out there would sign up in a heartbeat.

tbinmd 6y ago

So in my 23andme it shows me as having the T/T genotype of this marker. Does that mean I do or do not have the at risk marker?

boberto123 6y ago
I think you're good. From what I've read, the high risk is C. T/T is modern human. C is neanderthal. So C/C - youre 3 times more likely to get sick. C/T youre 1.5times more likely to get sick. Reference: https://www.gnxp.com/WordPress/2020/07/04/using-23andme-ancestry-family-tree-dna-to-identify-risk-allele-for-respiratory-failure-with-covid-19/comment-page-1/ https://www.gnxp.com/WordPress/2020/07/04/using-23andme-ance... === If this is wrong, someone correct me, but T/T results from 23andme seem to be 84.5% of the population (caucasion people), so YOU SHOULD BE FINE.

boberto123 6y ago

[Sars-Cov-2][Outcome risk] rs10490770 (Caucasisn only) Risk: AG, GG (~15.5% of population?) Safe: AA (~84.5% of population?) --- SO, if 23andme says its T/T and its actually A/A(as herf said), than 84.5% of the population would have T/T via 23andme. I am confused then. If I have T/T through 23andme, am I at higher risk when ~85% of the population has the same thing?

nigit 6y ago

Covid's no big deal cause I already got aids