You've been lied to about Globular Amphora Culture

[AndreiDNA]

You stated the reason I scored olive skin with your tool was because I am heterozygous for HERC2

yes, thats the biggest reason.
If you didnt have derived SLC24A5, SLC45A2, you would've scored light or dark brown. Out of genotypes that contribute to skin color, SLC45A2, SLC24A5, and OCA2/HERC2 had the biggest impact on you scoring olive skin.
Think of this. Spaniards, Italians, Turks, Georgians, they're olive skinned right? Well guess what. They have the same genotype as you. Practically every west eurasian aside from peninsular arabs has derived SLC45A2 and SLC24A5. The thing that sets them apart from northern europeans, skin color wise, is exactly OCA2/HERC2. There's other genes that contribute that i listed earlier, but HERC2/OCA2 is the biggest influencer in this regard. How is the tool supposed to know you're not a Spaniard? When your genotypes are typically spanish...

[Vessna]

yes, thats the biggest reason.
If you didnt have derived SLC24A5, SLC45A2, you would've scored light or dark brown. Out of genotypes that contribute to skin color, SLC45A2, SLC24A5, and OCA2/HERC2 had the biggest impact on you scoring olive skin.
Think of this. Spaniards, Italians, Turks, Georgians, they're olive skinned right? Well guess what. They have the same genotype as you. Practically every west eurasian aside from peninsular arabs has derived SLC45A2 and SLC24A5. The thing that sets them apart from northern europeans, skin color wise, is exactly OCA2/HERC2. There's other genes that contribute that i listed earlier, but HERC2/OCA2 is the biggest influencer in this regard. How is the tool supposed to know you're not a Spaniard? When your genotypes are typically spanish...

There are millions of pale skinned Europeans with brown eyes. If HERC2 heterozygosity or a brown eye genotype were enough to predict olive skin then a large portion of Eastern/Central/Northern Europeans would be olive skinned which obviously isn't the case. As I mentioned before, I do not have olive skin. I have to wear protective clothing and layers of sunscreen when I am outside. I might get a little color but burn over and over again if not protected. That's an objective finding, not some abstract. I posted images that you conveniently ignored.

[AndreiDNA]

There are millions of pale skinned Europeans with brown eyes. If HERC2 heterozygosity or a brown eye genotype were enough to predict olive skin then a large portion of Eastern/Central/Northern Europeans would be olive skinned which obviously isn't the case.

once again, you're ignoring the impact of other genes. A couple of posts above i showed globular amphora results with my trait predictor. Light skin, such as pale or white, is quite common in Europeans in combination with brown or hazel eyes. You're taking things to an extreme. I say "HERC2 contributes to skin color and is often the difference between white vs olive skin" and you take it to "Its impossible to have white skin if you carry ancestral HERC2".

[Vessna]

once again, you're ignoring the impact of other genes. A couple of posts above i showed globular amphora results with my trait predictor. Light skin, such as pale or white, is quite common in Europeans in combination with brown or hazel eyes. You're taking things to an extreme. I say "HERC2 contributes to skin color and is often the difference between white vs olive skin" and you take it to "Its impossible to have white skin if you carry ancestral HERC2".

But that is exactly what YOU are implying - your tool predicted over 95% probability of olive skin in my case because of HERC2 heterozygosity, despite me carrying only one ancestral allele. Make it make sense.

[AndreiDNA]

But that is exactly what YOU are implying - your tool predicted over 95% probability of olive skin in my case because of HERC2 heterozygosity, despite me carrying only one ancestral allele. Make it make sense.

Ok.
You have:
HERC2 - rs916977 1
^ the average northern european has 2 derived variants here
HERC2 - rs12913832 1
HERC2 - rs1129038 1
^these two are in a linked region, and some northern euros have heterozygous genotype, like you. most have 2 though. However in med countries, heterozygous genotype is the most common here
HERC2 - rs1667394 1
^The average European (including southern europeans) has 2 derived variants here.
OCA2 - rs4778138 : 1
^The average northern europeans has 2 derived variants here
So your HERC2 profile is pretty southern.

you also have:
ASIP - rs1015362 2
ASIP - rs4911414 0
ASIP - rs6142129 0

Which is quite "ancestral" or dark by average euro standards. This is another factor that predisposes you to olive, rather than white skin tone.

You do have this:
SLC45A2 - rs28777 2
SLC45A2 - rs16891982 2
SLC45A2 - rs28117 2
~SLC45A2 - rs4049255 2
~SLC45A2 - rs35401 2
SLC24A5 - rs1426654 2
Which prevents you from scoring stuff like light brown skin or dark brown skin tone.

Then you have this:
IRF4 - rs12203592 0
If you had derived variants here, like many NW europeans do, it would've led to a lighter skin shade prediction, increasing your odds of white skin. But you don't have any derived variants here.

Then you have:
KITLG - rs642742 2
KITLG - rs114873 2
KITLG - rs12821256 0

Which once again is the typical genotype for a Eurasian. Most folks from Spain to China have this genotype.

You also have:
TYRP1 - rs1408799 2
TYRP1 - rs683 2
TYRP1 - rs16929340 2
TYRP1 - rs1408794 2

Which is typical of NE europeans, but doesn't make a big impact on anything other than eye color.

Finally you have no derived MC1R variants, which could shift your skin color result to be white or palest, rather than Olive.

Basically you have all the variants for darker skin pigmentation in a European context, but have all the variants for lighter skin pigmentation in a West Eurasian/Eurasian context. Hence why the tool predicted you to have olive skin, like a southern european. It's what your genotypes suggest.

[Vessna]

yes, thats the biggest reason.
If you didnt have derived SLC24A5, SLC45A2, you would've scored light or dark brown. Out of genotypes that contribute to skin color, SLC45A2, SLC24A5, and OCA2/HERC2 had the biggest impact on you scoring olive skin.
Think of this. Spaniards, Italians, Turks, Georgians, they're olive skinned right? Well guess what. They have the same genotype as you. Practically every west eurasian aside from peninsular arabs has derived SLC45A2 and SLC24A5. The thing that sets them apart from northern europeans, skin color wise, is exactly OCA2/HERC2. There's other genes that contribute that i listed earlier, but HERC2/OCA2 is the biggest influencer in this regard. How is the tool supposed to know you're not a Spaniard? When your genotypes are typically spanish...

Maybe the issue is that the tool just isn't as accurate?
How is the tool supposed to know that an "olive skinned" blue eyed Spaniard isn't a Russian, or a blue eyed Turk isn't a Swede? Does nonancestral HERC2 makes them Northern European?

[AndreiDNA]

Maybe the issue is that the tool just isn't as accurate?
How is the tool supposed to know that an "olive skinned" blue eyed Spaniard isn't a Russian, or a blue eyed Turk isn't a Swede? Does nonancestral HERC2 makes them Northern European?

The point is to predict phenotype based on DNA, not on ethnicity. There are Spaniards with a nordic phenotype. The aim of my tool is so that these people get a Nordic phenotype prediction.

[Vessna]

Ok.
You have:
HERC2 - rs916977 1
^ the average northern european has 2 derived variants here
HERC2 - rs12913832 1
HERC2 - rs1129038 1
^these two are in a linked region, and some northern euros have heterozygous genotype, like you. most have 2 though. However in med countries, heterozygous genotype is the most common here
HERC2 - rs1667394 1
^The average European (including southern europeans) has 2 derived variants here.
OCA2 - rs4778138 : 1
^The average northern europeans has 2 derived variants here
So your HERC2 profile is pretty southern.

you also have:
ASIP - rs1015362 2
ASIP - rs4911414 0
ASIP - rs6142129 0

Which is quite "ancestral" or dark by average euro standards. This is another factor that predisposes you to olive, rather than white skin tone.

You do have this:
SLC45A2 - rs28777 2
SLC45A2 - rs16891982 2
SLC45A2 - rs28117 2
~SLC45A2 - rs4049255 2
~SLC45A2 - rs35401 2
SLC24A5 - rs1426654 2
Which prevents you from scoring stuff like light brown skin or dark brown skin tone.

Then you have this:
IRF4 - rs12203592 0
If you had derived variants here, like many NW europeans do, it would've led to a lighter skin shade prediction, increasing your odds of white skin. But you don't have any derived variants here.

Then you have:
KITLG - rs642742 2
KITLG - rs114873 2
KITLG - rs12821256 0

Which once again is the typical genotype for a Eurasian. Most folks from Spain to China have this genotype.

You also have:
TYRP1 - rs1408799 2
TYRP1 - rs683 2
TYRP1 - rs16929340 2
TYRP1 - rs1408794 2

Which is typical of NE europeans, but doesn't make a big impact on anything other than eye color.

Finally you have no derived MC1R variants, which could shift your skin color result to be white or palest, rather than Olive.

Basically you have all the variants for darker skin pigmentation in a European context, but have all the variants for lighter skin pigmentation in a West Eurasian/Eurasian context. Hence why the tool predicted you to have olive skin, like a southern european. It's what your genotypes suggest.

My issue isn't that the model predicted olive skin. It's the confidence level. If my actual phenotype doesn't match the prediction, maybe the model is overestimating the predictive value of those variants. I am curious to see what you have for my husband who has darker skin than me.

[AndreiDNA]

My issue isn't that the model predicted olive skin. It's the confidence level. If my actual phenotype doesn't match the prediction, maybe the model is overestimating the predictive value of those variants. I am curious to see what you have for my husband who has darker skin than me.

This confidence level is typical with bayesian likelyhood estimates. Have you ever used snipper3? try it, it will give you stuff like "80,000,000 more likely hazel than blue eyes" and such. It uses the same method as my tool's skin color estimate.
I modified the algorithm for eye and hair color, so that it wouldn't give out stuff like "99.5% odds of hazel eyes" for folks like you. I came up with my own mathematical method that still uses bayesian statistics but doesn't lead to a crazy confidence level.