New Extremely Accurate DNA Traits and Phenotype Predictor!

**Melkiirs** · 05-23-2024, 12:32 AM

Originally Posted by Beowulf

these?

Those are three IQ related SNPs. rs12202969 is another listed under bipolar instead.

**Beowulf** · 05-23-2024, 12:41 AM

Originally Posted by Melkiirs

What appears to be the case is that you are predicted so high for black hair that carrying a strong R variant (R142H) minimally elevated your red hair predication. For comparison I only carry one copy of the weak r variant V60L and am already at 0.592% for red hair. Given my other pigmentation variants and an MC1R with one copy of R142H and no V60L the predication would be 5.138%.

This is a very interesting case. Could you please share your skin pigmentation predication as well along with your genotype for the SLC45A2 variant rs16891982? If you are C;C for that variant you have a 7x greater chance of black hair as a European according to Snpedia. Hair color is polygenic but rs16891982 is very well studied so if you have the non-European genotype your prediction makes sense.

rs16891982 SNP doesn't appear for me

**Melkiirs** · 05-23-2024, 12:45 AM

Originally Posted by Beowulf

SLC45A2 - rs28777 1
SLC45A2 - rs16891982 999
SLC45A2 - rs28117 1
~SLC45A2 - rs4049255 1
~SLC45A2 - rs35401 1

Your rs16891982 SNP is not genotyped!!!??? Could you check your 23&me raw data directly for me? That is one of the most important pigmentation SNPs of all so the situation is very strange.

**Beowulf** · 05-23-2024, 12:45 AM

Originally Posted by Melkiirs

Your rs16891982 SNP is not genotyped!!!??? Could you check your 23&me raw data directly for me? That is one of the most pigmentation SNPs of all so the situation is very strange.

i don't have 23 and Me

**Melkiirs** · 05-23-2024, 12:47 AM

Originally Posted by Beowulf

i don't have 23 and Me

Where does your raw data come from then? I thought you tested with 23&me?

**Beowulf** · 05-23-2024, 12:47 AM

Originally Posted by Melkiirs

Where does your raw data come from then? I thought you tested with 23&me?

FTDNA

**Melkiirs** · 05-23-2024, 12:52 AM

Originally Posted by Beowulf

FTDNA

My mistake. I'm shocked FamilyTreeDNA did not genotype that SNP. Could you check you raw data file? I might have to reach out to Andrei about issues with FamilyTreeDNA raw data.

**Beowulf** · 05-23-2024, 12:53 AM

Originally Posted by Melkiirs

My mistake. I'm shocked FamilyTreeDNA did not genotype that SNP. Could you check you raw data file? I might have to reach out to Andrei about issues with FamilyTreeDNA raw data.

I did have checked my raw data and yeah it isn't genotypized.

**Melkiirs** · 05-23-2024, 01:04 AM

Originally Posted by Beowulf

I did have checked my raw data and yeah it isn't genotypized.

OK, it might have just been a normal error in an unlucky position. With AncestryDNA raw data SNPs that were not determined use a "0" instead of a base letter and FTDNA should use a similar placeholder. For example from my file:

Change rs16891982 to G;G which is the strong majority across Europe and then maybe your phenotype results will be much more accurate.

**calxpal** · 05-23-2024, 01:38 AM

Hi I've tried this tool as well, it's a neat tool but I have to be honest some of my trait predictions on it were super far off from reality especially some of the phenotypes it assigned me in the oracle. I still like the tool though. The eye color prediction for me was very good though. The ancestry report for me was unfortunately extremely inaccurate, but it was neat to try. I hope the tool gets some good updates.