AI portrait
This portrait was algorithmically built from this dog's genome: their genotype at 8 morphology loci (coat length, curl, color, ear set, body size, head shape, skull, furnishings) plus their position within the 14,478-dog atlas. The same dog always reproduces the same portrait. A different dog with different alleles gets a different portrait.
PILOT_grlsCOEDU1TT
PILOT_grlsCOEDU1TT is a Golden Retriever from the GRLS research cohort. One of 14,478 dogs who built the atlas.
See PILOT_grlsCOEDU1TT in the atlasPILOT_grlsCOEDU1TT sits well inside the Golden Retriever cluster - a fairly typical example.
- Predicted large by the six body-size genes the atlas reads (IGF1, HMGA2, SMAD2, LCORL, STC2, ADAMTS17).
- Standard leg length. No chondrodysplasia retrogene variant.
- Wire-coat furnishings - the eyebrow + beard variant at RSPO2.
The five dogs in the atlas whose genomes sit closest to PILOT_grlsCOEDU1TT's. Click any of them to keep exploring.
PILOT_grlsCOEDU1TT sits in the Golden Retriever cluster, with genome overlap to Labrador Retriever - sister breeds nearby in the atlas.
Breed similarity from non-negative least squares against 91 breed centroids in PCA-256 space, corrected for atlas sample-size imbalance. Without correction, Goldens (22% of the atlas) leak into every dog's raw NNLS breakdown; with it, the bias falls out. Raw fractions stay in the dataset for re-derivation. Methodology.
- Golden Retriever 80%
- Labrador Retriever 20%
From the Golden Retriever Lifetime Study . Breed-page reference: Golden Retriever.
Full genotype detail click to expand
The actual allele call at each locus's representative SNP for this dog. Each gene name links to its page where you can see the per-breed frequency table and the direction of effect.
Technical details click to expand
The numbers behind the placement. Useful for researchers reproducing the math or debugging an unexpected position; not interesting to most readers.
y 9.992
z 8.220
The 3 PCs on which PILOT_grlsCOEDU1TT scores most extreme, with the 3 highest-loading SNPs on each. Foundation for the future genome-ring visualization.
- chr19:19,197,831 loading 0.0370
- chr1:99,774,502 loading -0.0342
- chr4:60,708,893 loading 0.0329
- chr3:6,234,729 loading 0.0318
- chr23:19,305,339 loading 0.0317
- chr19:17,376,904 loading 0.0316
- chr12:606,598 loading 0.0433
- chr12:4,288,226 loading -0.0385
- chr12:1,150,085 loading 0.0383