The Unz Review - Mobile
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 iSteve BlogTeasers
Can Your Genes Predict Your Looks?
Email This Page to Someone

 Remember My Information



=>
Search Text Case Sensitive  Exact Words  Include Comments

From MIT Technology Review:

Rewriting Life

Does Your Genome Predict Your Face? Not Quite Yet

Genomics pioneer J. Craig Venter says if he had your genome he could pick you out of a crowd.

by Antonio Regalado September 7, 2017

On Monday, the California gene-hunting company Human Longevity published a paper making the bold claim that it can identify individuals using their genomes to predict what their faces looks like.

The assertion—that your DNA can be used to create a photo-like reconstruction of you—has potentially big implications. It would allow police to pick suspects out of a lineup using a blood spot and it would mean no genome collected for research is truly private.

But a withering reaction to the face-prediction paper by scientists on social media is probably not what Human Longevity’s founder, the famed genomics expert J. Craig Venter, had in mind.

Venter was an entrepreneur who elbowed his way into the government’s Human Genome Project in the 1990s and vastly sped it up.

According to two experts who reviewed the paper—and one former employee—Venter can’t actually pick a person out of a crowd using a genome, and his report had difficulty finding a publisher.

“Craig Venter cannot predict faces,” Yaniv Erlich, the chief scientific officer of MyHeritage.com, a genealogy website, said bluntly on Twitter. To make the point, Erlich posted Venter’s prediction of his own face, noting that it looked more like actor Bradley Cooper than the biologist-turned-entrepreneur.

From L to R: Craig Venter, Venter’s prediction of what he looks like based on his own DNA, Bradley Cooper

… But skeptics say Human Longevity actually uses a person’s race and sex—easily measured from simpler DNA tests and not a new idea—to create pictures that portray average faces, not specific ones, as the company said.

Ironically, Venter played a large role in promoting the popular myth that Race Does Not Exist with talking points he asserted at the White House Rose Garden ceremony in 2000, such as “Race is a social concept, not a scientific one.”

Using its method, the company reported it could pick the right person from a lineup of 20 photographs about 70 percent of the time. But after discarding people of a different sex and race than the subject, accuracy dropped drastically. Used to pick a specific European man out of a lineup of 20 other European men, it was right 11 percent of the time.

So that’s a little over twice random luck (5%). Using advanced genetics to identify facial features gets you from 5% to 11%, while using basic genetics to identify sex and race gets you from 11% to 70%.

It’s almost as if the Bill Clinton-Venter-Collins speeches in the Rose Garden in 2000 about how genetics proves race isn’t scientific were political rather than scientific.

 
Commenters to Ignore...to FollowEndorsed Only
    []
  1. TangoMan says:

    Jerry Pournelle has passed.

    Read More
    • Replies: @Romanian
    I am so sorry to hear that. I really liked his work, though I did not keep up with his blogging.
    ReplyAgree/Disagree/Etc.
    AgreeDisagreeLOLTroll
    These buttons register your public Agreement, Disagreement, Troll, or LOL with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used once per hour.
    Sharing Comment via Twitter
    /isteve/can-your-genes-predict-your-looks/#comment-1999057
    More... This Commenter This Thread Hide Thread Display All Comments
  2. AB- says:

    I look just like my Dad. Anybody that knows both of us tells me. Kid lived 2 doors down, same thing.

    Read More
    • Replies: @Buzz Mohawk

    I look just like my Dad. Anybody that knows both of us tells me.
     
    Indeed, and the older we get, the more we look, act and live like our parents. It's downright annoying. It exposes all those lies about how unique we are and how we can be anything we want to be. Puts them right up there with Santa Claus.
    , @Bill Jones
    There's a picture of me at 16 floating around- the current age of my boy and I've had 5 people say : Well, there's no doubting who the father is.
    , @Anon
    Does your mom know about the kid down the street?
    , @Neil Templeton
    Me too. People always said I looked just like my dad. He died at 48 of lung cancer. Two packs a day. I'm 56 now, 1/2 pack and still smokin'! I laugh in the face of genetics, the faith of the observant.
    , @Anonymous
    At several of the places where my dad worked, several years later someone had a kid that looked sort of like him. Probably coincidental.
  3. njguy73 says:

    As Mets announcer Ralph Kiner said of some ballplayer whose father also played in the majors, “There’s a lot of heredity in that family.”

    Read More
  4. @AB-
    I look just like my Dad. Anybody that knows both of us tells me. Kid lived 2 doors down, same thing.

    I look just like my Dad. Anybody that knows both of us tells me.

    Indeed, and the older we get, the more we look, act and live like our parents. It’s downright annoying. It exposes all those lies about how unique we are and how we can be anything we want to be. Puts them right up there with Santa Claus.

    Read More
  5. @AB-
    I look just like my Dad. Anybody that knows both of us tells me. Kid lived 2 doors down, same thing.

    There’s a picture of me at 16 floating around- the current age of my boy and I’ve had 5 people say : Well, there’s no doubting who the father is.

    Read More
  6. eD says:

    If genes can’t predict your looks, what can they predict?

    Read More
    • Replies: @Lot
    With near 100% accuracy: ear lobe type and ear wax type
  7. Lot says:

    Nobody took up my suggestion to plug the IQ genes in the new study into their 23andme data to create a predicted IQ for themselves and compare with actual.

    It would probably end up being from 97-103 for almost everyone. We still only have a fraction of them identified.

    I did it with a few, and about half were included in the 23andme sequence. They only sequenced a small fraction of my genes, but did a good job targeting the interesting ones.

    Regarding appearance, the predictions for my coloration and hair type were better than random but still somewhat off.

    Read More
    • Replies: @Sunbeam
    "Nobody took up my suggestion to plug the IQ genes in the new study into their 23andme data to create a predicted IQ for themselves and compare with actual. "

    That brings to mind a question: How thorough (for want of a better word) is what 23andme does? Or any of these commercial dna analysis outfits?

    Is it as in depth as what someone in a lab with some kind of Neanderthal dna would do? I somehow have the idea that 23andme doesn't spend as much time running their test to be as accurate as some other approaches.

    We got a pretty educated crowd here, so any enlightenment available? My take was that it would be too expensive to do as accurate a job as state of the art allows now for something that perform as many times as they do for the price.

    Meaning it might not be amenable to the kinds of things like IQ and appearance prediction.
  8. I actually don’t think the prediction for Venter is that bad, assuming that the prediction is of like a 30 year old version of the face. The basic facial structure is not terrible, the main differences are the facial hair and the aging.

    Read More
    • Replies: @Chrisnonymous
    That was my initial reaction too. Just based on those photos, the main difference between Venter and the prediction is that Venter doesn't have such a convex, downward-hooking nose. (The tip of the nose really does look more like the actor.) Maybe his eye sockets are a bit deeper too, but overall very similar!

    So I looked for a photo of a younger Venter. For example:
    http://images.the-scientist.com/content/figures/images/yr1998/jan/jan_art/venter.jpg

    If Venter and the actor both look like the prediction, they should look like each other too... Does the younger Venter look like Cooper?
    https://www.famousbirthdays.com/headshots/bradley-cooper-6.jpg

    Seems like no, but the images are hard to compare as Venter is obviously out of shape compared to Cooper.
  9. Anon says: • Disclaimer
    @AB-
    I look just like my Dad. Anybody that knows both of us tells me. Kid lived 2 doors down, same thing.

    Does your mom know about the kid down the street?

    Read More
  10. To summarize the main point of the cited article: It will take some time before we find all the genes that affect facial appearance (and adequately understand how they work). Only then will DNA facial reconstruction come into its own.

    That always seemed like a reasonable position to me, but you’ll hear claims that we’re practically there already. The buzz tends to come from the Big Data crowd, glorified IT people who ignore the limitations of what they’re doing. Faster computers and slightly better algorithms will never substitute for gene identification studies and the like. Other companies like Parabon Nanolabs will claim better accuracy, but where’s the published research to back it up?

    Non-rhetorical question for Steve and his readers. How long before we actually get there? I’m thinking at least a decade, and that feels very optimistic.

    Read More
  11. Carneades says:

    When I ran the eye color prediction program on GEDmatch it came back with a photo that was almost spot on, unlike 23andme which didn’t even get the color right.

    Read More
  12. Carneades says:

    When I ran the eye color prediction program for my DNA sample on GEDmatch it came back with a photo of an iris that was almost spot on, unlike 23andme which didn’t even get the color right.

    Read More
  13. @AB-
    I look just like my Dad. Anybody that knows both of us tells me. Kid lived 2 doors down, same thing.

    Me too. People always said I looked just like my dad. He died at 48 of lung cancer. Two packs a day. I’m 56 now, 1/2 pack and still smokin’! I laugh in the face of genetics, the faith of the observant.

    Read More
  14. Brings back memories of this. Imagine, there are people who are opposed to using DNA to identify the race of a rapist or murderer.

    https://www.wired.com/2007/12/ps-dna/

    Read More
  15. Anonymous says: • Disclaimer
    @AB-
    I look just like my Dad. Anybody that knows both of us tells me. Kid lived 2 doors down, same thing.

    At several of the places where my dad worked, several years later someone had a kid that looked sort of like him. Probably coincidental.

    Read More
    • Replies: @Curle
    People say this fellow, Paul Cochrane, looks like Paul McCartney. Funny that his mother was in a relationship with the Beatle the month he was conceived.










    https://groupieblog.files.wordpress.com/2011/03/philip-howarth-cochrane.jpg
  16. such as “Race is a social concept, not a scientific one.”

    Is there such a thing as a concept that isn’t social?

    Read More
  17. @415 reasons
    I actually don't think the prediction for Venter is that bad, assuming that the prediction is of like a 30 year old version of the face. The basic facial structure is not terrible, the main differences are the facial hair and the aging.

    That was my initial reaction too. Just based on those photos, the main difference between Venter and the prediction is that Venter doesn’t have such a convex, downward-hooking nose. (The tip of the nose really does look more like the actor.) Maybe his eye sockets are a bit deeper too, but overall very similar!

    So I looked for a photo of a younger Venter. For example:

    If Venter and the actor both look like the prediction, they should look like each other too… Does the younger Venter look like Cooper?

    Seems like no, but the images are hard to compare as Venter is obviously out of shape compared to Cooper.

    Read More
    • Replies: @Cortes
    Venter looks like Der Kaiser (the football/soccer one):

    https://en.m.wikipedia.org/wiki/Franz_Beckenbauer
  18. Lot says:
    @eD
    If genes can't predict your looks, what can they predict?

    With near 100% accuracy: ear lobe type and ear wax type

    Read More
  19. I don’t quite get this. Do twins raised apart look different? If not, how could genes not predict appearance?

    Read More
    • Replies: @unzerker
    Exactly! Anybody who has seen identical twins knows that genes predict your looks by almost 100%.
    , @Altai
    The loci involved are too numerous and not fully known. Additional environmental factors like the environment of the womb can also have an effect on hormone levels during development but it's not known how significant or how significant hormone levels during development are to facial development.

    The issue was Venter didn't have the information or tools to do it, not that it couldn't be done in theory. His paper did not show what he was implying it had shown, it was nothing new or interesting. There are a few projects going out to find the loci most responsible for facial features.

    , @Jim
    Every picture of identical twins I've seen have looked very similar. I thought that was why they were called "identical twins".
  20. Hugh says:

    This seems a step too far at the current state of play. Even identical twins develop different faces as they age: nurture is clearly playing a part here.

    It would be better for Venter to tell us how tall someone is just by looking at his genes.

    Read More
    • Replies: @Opinionator
    Other cellular components are not entirely the same.
  21. Romanian says: • Website
    @TangoMan
    Jerry Pournelle has passed.

    I am so sorry to hear that. I really liked his work, though I did not keep up with his blogging.

    Read More
  22. unzerker says:
    @Chrisnonymous
    I don't quite get this. Do twins raised apart look different? If not, how could genes not predict appearance?

    Exactly! Anybody who has seen identical twins knows that genes predict your looks by almost 100%.

    Read More
  23. Sunbeam says:
    @Lot
    Nobody took up my suggestion to plug the IQ genes in the new study into their 23andme data to create a predicted IQ for themselves and compare with actual.

    It would probably end up being from 97-103 for almost everyone. We still only have a fraction of them identified.

    I did it with a few, and about half were included in the 23andme sequence. They only sequenced a small fraction of my genes, but did a good job targeting the interesting ones.

    Regarding appearance, the predictions for my coloration and hair type were better than random but still somewhat off.

    “Nobody took up my suggestion to plug the IQ genes in the new study into their 23andme data to create a predicted IQ for themselves and compare with actual. ”

    That brings to mind a question: How thorough (for want of a better word) is what 23andme does? Or any of these commercial dna analysis outfits?

    Is it as in depth as what someone in a lab with some kind of Neanderthal dna would do? I somehow have the idea that 23andme doesn’t spend as much time running their test to be as accurate as some other approaches.

    We got a pretty educated crowd here, so any enlightenment available? My take was that it would be too expensive to do as accurate a job as state of the art allows now for something that perform as many times as they do for the price.

    Meaning it might not be amenable to the kinds of things like IQ and appearance prediction.

    Read More
    • Replies: @Lot
    It is very accurate. Error rates have been very low for a long time, the sequencing machines more compete on price and speed.

    Your choices now are to spend $210 on 23andme's partial sequence and very useful and user friendly interface or a full sequence which I think coats about $1500 now. Razib's archive on this site has a lot more info.

    I hope more people sign up, and also opt in to as many studies and answer as many site questions as possible to further this research. It is interesting and useful stuff on a lot of levels, but you are also helping medical research in a much bigger way than donations to charity.
  24. Altai says:
    @Chrisnonymous
    I don't quite get this. Do twins raised apart look different? If not, how could genes not predict appearance?

    The loci involved are too numerous and not fully known. Additional environmental factors like the environment of the womb can also have an effect on hormone levels during development but it’s not known how significant or how significant hormone levels during development are to facial development.

    The issue was Venter didn’t have the information or tools to do it, not that it couldn’t be done in theory. His paper did not show what he was implying it had shown, it was nothing new or interesting. There are a few projects going out to find the loci most responsible for facial features.

    Read More
    • Replies: @Chrisnonymous
    Yeah, forgot about pre-natal environment (blush), but if all the encoding contributors haven't been identified, it mostly means Venter's program has to revised, not that it's wrong-headed.
  25. 5371 says:

    I bet there was a small sample and the technique is actually no better than chance.
    Venter has always been a conman and fame deserted him several years ago.

    Read More
  26. anonymous says: • Disclaimer

    Bradley Cooper and Craig Venter share at least one attribute. Look at Cooper’s eyes and sockets. Those indicate someone who wears, or wore, glasses. There is a definite look of that that escapes the best makeup artists.

    Read More
  27. Cortes says:
    @Chrisnonymous
    That was my initial reaction too. Just based on those photos, the main difference between Venter and the prediction is that Venter doesn't have such a convex, downward-hooking nose. (The tip of the nose really does look more like the actor.) Maybe his eye sockets are a bit deeper too, but overall very similar!

    So I looked for a photo of a younger Venter. For example:
    http://images.the-scientist.com/content/figures/images/yr1998/jan/jan_art/venter.jpg

    If Venter and the actor both look like the prediction, they should look like each other too... Does the younger Venter look like Cooper?
    https://www.famousbirthdays.com/headshots/bradley-cooper-6.jpg

    Seems like no, but the images are hard to compare as Venter is obviously out of shape compared to Cooper.

    Venter looks like Der Kaiser (the football/soccer one):

    https://en.m.wikipedia.org/wiki/Franz_Beckenbauer

    Read More
  28. Curle says:
    @Anonymous
    At several of the places where my dad worked, several years later someone had a kid that looked sort of like him. Probably coincidental.

    People say this fellow, Paul Cochrane, looks like Paul McCartney. Funny that his mother was in a relationship with the Beatle the month he was conceived.

    Read More
    • Replies: @Cortes
    Looks like Patrick Roberts...

    https://en.m.wikipedia.org/wiki/Patrick_Roberts
  29. Jim says:
    @Chrisnonymous
    I don't quite get this. Do twins raised apart look different? If not, how could genes not predict appearance?

    Every picture of identical twins I’ve seen have looked very similar. I thought that was why they were called “identical twins”.

    Read More
  30. @Hugh
    This seems a step too far at the current state of play. Even identical twins develop different faces as they age: nurture is clearly playing a part here.

    It would be better for Venter to tell us how tall someone is just by looking at his genes.

    Other cellular components are not entirely the same.

    Read More
  31. Lot says:
    @Sunbeam
    "Nobody took up my suggestion to plug the IQ genes in the new study into their 23andme data to create a predicted IQ for themselves and compare with actual. "

    That brings to mind a question: How thorough (for want of a better word) is what 23andme does? Or any of these commercial dna analysis outfits?

    Is it as in depth as what someone in a lab with some kind of Neanderthal dna would do? I somehow have the idea that 23andme doesn't spend as much time running their test to be as accurate as some other approaches.

    We got a pretty educated crowd here, so any enlightenment available? My take was that it would be too expensive to do as accurate a job as state of the art allows now for something that perform as many times as they do for the price.

    Meaning it might not be amenable to the kinds of things like IQ and appearance prediction.

    It is very accurate. Error rates have been very low for a long time, the sequencing machines more compete on price and speed.

    Your choices now are to spend $210 on 23andme’s partial sequence and very useful and user friendly interface or a full sequence which I think coats about $1500 now. Razib’s archive on this site has a lot more info.

    I hope more people sign up, and also opt in to as many studies and answer as many site questions as possible to further this research. It is interesting and useful stuff on a lot of levels, but you are also helping medical research in a much bigger way than donations to charity.

    Read More
  32. Cortes says:
    @Curle
    People say this fellow, Paul Cochrane, looks like Paul McCartney. Funny that his mother was in a relationship with the Beatle the month he was conceived.










    https://groupieblog.files.wordpress.com/2011/03/philip-howarth-cochrane.jpg
    Read More
  33. @Altai
    The loci involved are too numerous and not fully known. Additional environmental factors like the environment of the womb can also have an effect on hormone levels during development but it's not known how significant or how significant hormone levels during development are to facial development.

    The issue was Venter didn't have the information or tools to do it, not that it couldn't be done in theory. His paper did not show what he was implying it had shown, it was nothing new or interesting. There are a few projects going out to find the loci most responsible for facial features.

    Yeah, forgot about pre-natal environment (blush), but if all the encoding contributors haven’t been identified, it mostly means Venter’s program has to revised, not that it’s wrong-headed.

    Read More
Current Commenter says:

Leave a Reply - Comments are moderated by iSteve, at whim.


 Remember My InformationWhy?
 Email Replies to my Comment
Submitted comments become the property of The Unz Review and may be republished elsewhere at the sole discretion of the latter
Subscribe to This Comment Thread via RSS Subscribe to All Steve Sailer Comments via RSS
PastClassics
A simple remedy for income stagnation
Confederate Flag Day, State Capitol, Raleigh, N.C. -- March 3, 2007
The major media overlooked Communist spies and Madoff’s fraud. What are they missing today?
Are elite university admissions based on meritocracy and diversity as claimed?
The evidence is clear — but often ignored