Skip to content

Technical Details

June 30, 2013

This is the first in a series of technical posts I hope to put together over the coming weeks about the process of creating Stranger Visions. Starting from the lab end I will describe in detail how I create these portraits from DNA.

The first step of course is extracting DNA from a sample, that can be done with a kit readily obtainable from Qiagen, the “DNA Investigator Kit”.

After I extract the DNA* I do a test PCR for a specific alu insertion using a primer (full kit is also available) from Carolina. This gives me an idea of whether I successfully extracted chromosomal DNA or not. If the results look bad on this there is probably no reason to go further with this sample.

Next is looking at maternal ancestry from mitochondrial DNA. A couple different primer/kits are also available from carolina to look at this region amplifying for sequencing and  haplotyping.

If all these results look good I continue to look at 3 SNPs associated with particular traits and 1 region on the y chromsome:

rs12913832 – for eye color

SRY gene – present on Y chromosome, I use to indicate gender

rs9939609 – as one of many indicators of obesity

rs4648379 – preliminary research, correlated with the size of the nose

I have been using a free tool on IDT for designing primers to target these regions using PCR. If the result of the reactions looks good in a gel I send them for sequencing.
So those are SNPs I have had success so far analyzing in the lab myself, but there are a whole lot more clues I have on a “master list”, the things I can find out from a 23andme profile for example. In addition to SNPs related to specific traits this includes Y chromosome haplogroup information from males to get an idea of paternal ancestry, as well as Ancestry Informative Markers, specific SNPs that are correlated with population studies giving an overall picture of ancestry beyond the direct maternal and paternal lineage. I am pasting my master list here below (excluding the ancestry info I just mentioned and linked to above as that would increase the list a thousand fold). Also, here is a link to appearance related SNPs on snpedia.

*I have recently discovered that whole genome amplification can also be performed after the extraction step and isn’t as expensive as I previously though – I will likely pursue this in future experiments.

SNP allele associated trait
rs642742 AA west african
rs1426654 AA light skinned european
rs2814778 CC malaria resistance
rs17646946  GG curlier hair in europeans
rs11803731 AA straighter hair
rs12896399 TT blue eyes vs green if not brown – prelim research
rs1426654 AA euro ancestry – light skinned
rs12913832 GG blue eyes
rs1667394 CC decreased odds blond hair
rs1805007 TT higher odds red hair
rs17822931  TT dry earwax
rs1851665 AA greater chance of obesity
rs6625163 AA baldness
rs2153271 TT more freckling
rs7349332 CT;TT curlier hair
rs6732426 CC curlier hair
rs3827760  AG thinner hair (asian)
rs9939609 AA higher BMI
AT typical BMI
TT lower BMI
rs6548238   CC typical BMI
CT;TT lower BMI
rs925946  GT typical BMI
TT higher BMI
GG lower BMI
rs7138803   AG typical BMI
AA higher BMI
GG lower BMI
rs13130484  CT typical BMI
TT higher BMI
CC lower BMI
rs4788102  AG typical BMI
AA higher BMI
GG lower BMI
rs10838738 AG typical BMI
GG higher BMI
AA lower BMI
rs10871777  GG;AG higher BMI
rs6113491 AA baldness
rs6152 GG baldness
rs1385699 TT baldness
rs1393350 AA blue eyes rather than green IF non-brown – preliminary research
rs1533995 AA more pronounced crypts
AG typical
GG less pronounced crypts
rs3739070 AA typical
AC,CC less iris furrows
rs12896399 GG more pronounced pigementation rings
TT absent pigmentation rings
CT slightly higher odds red hair
rs1805009 CC red hair
TT increased odds brown hair
CT typical
CC less freckling
rs4778138 GG less freckling
AA more freckling
rs1805007 CT;TT more freckling
rs1042602 AA absence of freckles
rs2733832 CC darker skin
CT;TT lighter skin
rs1800414 C each c -> 1.3 melanin unit lighter skin in east asians
rs16891982 GG european light skinned
preliminary facial research source
rs4648379 C/T edge of nostril -> tip of lower nose
rs974448 C/T mid eye to top of nose
rs17447439 A/G distance between eyes
rs6555969 C/T distance between cheekbone and nose, ALSO distance between eye and top nose
rs805722 C/T distance between eye and top of nose
Eye and Skin color source
  1. Dave permalink
    July 18, 2013 2:17 am

    Do you think that the ‘information’ and ‘information process’ that you are helping develop might be used in a non artistic way?

    It wasn’t long after the concept of ‘evolution’ was presented that the idea of ‘eugenics’ caused a great deal of harm.

    Just saying that developing a new ‘art’ through technology needs a bit of caution, especially when we can pinpoint exact times and locations.

    Though I have to say that what you are doing in the way of a good crime story is quite nice.

    Just think,

    HDH. Scientific Investigation…


    • September 5, 2013 9:23 pm

      I am hopeful that my work will call attention to the potential future issues so we can confront them early on.

  2. Laura permalink
    September 5, 2013 12:13 am

    This is super fascinating! Do you take commissions? I would love to see what my DNA portrait looks like.

  3. Mickayel Mkrtchian permalink
    September 5, 2013 1:11 am

    I was on facebook and happened to see a missing girl posting by her mother. I clicked on the link, which lead me to the missing and exploited children’s page. To my horror, the page was full of childrens picture, some current, some from the past, missing by abduction or runaway.
    I wondere where these kids are now, if they even are alive.
    They seem to go into a vaccum not to be seen again.
    Then I thought the possible ways of finding them.
    One thing that cam to mind was DNA, either from live individuals or people unknown, possibley even dead.
    So I seached google for DNA face and came to your site. Apparently there is a good method of developing a face or facial variations by pure DNA. Does the technique have an age parameter, what age face you display?
    If this is the case, a numerical scoring algorithm can be set up, where the missin childrens pictures are scaned and given a set of values, and any and all available known (if the permission of the onwers are given) or unknown DNA samples are analysed, and instead of a physical face, a similar se of values are generated. Then these values are matched with some scoring system.
    This would give you a face to DNA match, and help in finding what happened to these kids.
    I don’t know if a curren system like this exists, but if it doesn’t what can be do e in getting it developed, in technical aspect.
    Anything to help shed any light on a missing child case would be amazing to the parents of these children.

  4. Tina permalink
    September 5, 2013 9:37 am

    Interesting technique! The CNN article states: “Once she obtains the sequencing information, she takes the traits she’s gathered from the individual’s DNA and feeds it into a computer program to generate a 3-D model of a face.”

    Can you share what the “computer program” is that generates this 3D model? Is it PrimerQuest or something else?

    Fun stuff.

  5. October 16, 2013 4:16 pm

    What a great blog! Makes me believe that one can Extract Art from Science! You are doing wonderful things with your strangervisions Heather!


  1. Extracting Art from Science | Outside The Lines
  2. No podemos vivir sin tocar el mundo | Desconexión Económica
  3. DNA at Crime Scene can be used to Create Offender’s Mugshot | GoSouthOnline
  4. DNA at Crime Scene can be used to Create Offender's Mugshot - Naija247news

Comments are closed.

%d bloggers like this: