What My Genome Actually Says

The Error That Produced Wrong Results — and How It Was Fixed

The initial automated analysis used 55 pre-selected ancestry-informative markers (AIMs). Every single one came back as "homozygous reference" — producing the absurd result of 0% European ancestry for someone with documented Iberian descent and a European maternal lineage. This was not a biological finding; it was a software artefact.

The fix: merge the full WGS against 2,513 samples from 1000 Genomes Phase 3 using bcftools, extracting dosages at 58,185 shared biallelic SNPs. At these positions, genotypes are emitted regardless of whether they match the reference. With 58,114 non-reference genotype calls confirmed, the corrected maximum-likelihood ancestry estimation runs on real data.

Left (red border): the artefactual first-round analysis — 0% European, which is impossible. Centre: corrected MLE raw weights using IBS (Iberian), PEL (Peruvian), and YRI (Yoruba) proxy populations. Right: comparison of MLE and PCA k-nearest-neighbour methods, converging on the same estimate.

Where I Come From: The Corrected Picture

Two independent methods — maximum-likelihood estimation against 1000 Genomes proxies, and principal component analysis with k-nearest neighbours — converge on the same result: predominantly Iberian European ancestry, with a significant Native American component and a meaningful African contribution.

The PCA k-nearest-neighbour result (k=20) places the genome at: PEL 55%, MXL 30%, CLM 15% — consistent with the MLE. In PC1–PC2 space the genome sits equidistant from the European centroid (distance 0.096) and the African centroid (0.097), confirming substantial Iberian ancestry plus an Afro-Colombian contribution. PC3 rises above all European populations, reflecting Native American ancestry.

PCA (PC1–PC2 and PC1–PC3) placing the genome (★) against 2,500+ 1000 Genomes samples. PC2 tracks the European–African axis; PC3 tracks Native American ancestry. The genome sits between the Iberian cluster and the Latino/admixed cluster.

My Father's Line: R1b-DF27* — Bronze Age Iberia

The Y chromosome is inherited exclusively from father to son, unchanged except for occasional mutations. These mutations allow geneticists to trace paternal lineages across millennia. My Y chromosome belongs to R1b-DF27* — confirmed directly from the VCF at the relevant SNP positions.

R1b-DF27 splits into two sub-clades: Z195+ (Basque / Eastern Iberian arc) and DF27* (xZ195) (basal paragroup in western Iberia — Portugal, Galicia). The absence of Z195 places this lineage in the latter — concentrated at 35–40% frequency in Portugal and Atlantic-facing Spain, and at ~38% in Colombian mestizo males through the colonial male bottleneck.

Paternal lineage analysis from the corrected report. M343 directly detected in chrY.vcf. Z195 absent → basal DF27* paragroup. Note: Y chromosome coverage was ~45.8× effective depth — low due to repetitive regions — but the key diagnostic SNPs were callable.

Colombia's ~38% R1b-DF27* frequency is one of the highest outside Iberia and directly traces to the Iberian colonial male bottleneck — the fact that Spanish male colonisers outnumbered female colonisers dramatically, making their Y chromosomes disproportionately common in today's Colombian male population.

My Mother's Line: H3/H3g — Ice Age Survivors

Mitochondrial DNA is passed exclusively from mother to child, tracing the maternal line back in an unbroken chain. The mitochondrial genome was sequenced at an average depth of 3,662× — over 100× more coverage than the nuclear genome — giving very high confidence in haplogroup assignment.

My maternal haplogroup is H3/H3g, assigned from the following markers (all confirmed at allele frequency ≥ 95%):

H1 is excluded — the diagnostic 3010 G→A variant is absent. The HVR1 constellation (16111T · 16187T · 16223T · 16290T · 16362C) is documented in H3g and closely related western Iberian H3 branches, confirming the sub-haplogroup assignment.

Maternal lineage analysis. H backbone confirmed at 3,662× depth. H1 excluded by absent 3010 G>A diagnostic. H3 frequency map shows highest concentration in Portugal, Basque Country, Berber North Africa, and the Canary Islands.

The canonical Iberian pattern: steppe males (R1b, Bell Beaker) arrived ~2500 BCE and replaced nearly all prior Y-lineages, but Mesolithic women persisted — H3/H3g is a pre-Bell Beaker European maternal lineage, unbroken for over 10,000 years. The genome carries both halves of this story.

The Jewish Ancestry: What the Genome Can and Cannot Detect

My family holds a Portuguese certificate of Sephardic Jewish origin (Comunidade Israelita de Lisboa, 2021), documenting descent from Jews expelled from Iberia in 1492–1497 who converted to Christianity and eventually settled in New Granada (modern Colombia).

The genomic picture is nuanced. In PCA space, the genome positions closer to Sephardic Jewish reference samples than to the average Iberian — consistent with Converso ancestry. At the same time, no statistically distinguishable Middle Eastern signal exists above the background of Iberian ancestry. These are not contradictory findings.

The PCA distance analysis confirms: the genome is ~3–4× closer to Iberians (IBS: 0.038) than to HGDP Middle Eastern populations (Druze: 0.147, Palestinian: 0.143, Bedouin: 0.151, Mozabite: 0.138). The Sephardic proximity visible in the PCA reflects shared Iberian ancestry, not a distinct Middle Eastern component.

PCA including Sephardic Jewish reference samples (Behar et al. 2010 dataset). The genome (★ JUAN) clusters in a region intermediate between Iberians and Sephardic Jews — reflecting the deep Iberian ancestry shared between Converso families and Old Christian Iberians after 500 years of admixture.

Euclidean PCA distances from the genome to reference populations. Iberians (IBS) are the closest group. Middle Eastern populations are 3–4× farther. The interpretation panel (right) explains why this does not contradict the certificate.

Educational Attainment Genetics — and the EA/Cognition Gap

Three independent GWAS datasets were used. All three place the educational attainment (EA) score substantially above average. The most rigorous uses Okbay et al. 2022 (Nature Genetics, N~3 million participants):

The most consistent finding across all three studies is the gap between the EA score and the cognitive proxy — roughly 1.0–1.4 standard deviations. Within the Okbay 2022 framework, EA genetic variance decomposes into ~65% cognitive and ~35% non-cognitive components (conscientiousness, persistence, SES-correlated traits). An EA score that substantially exceeds the cognitive proxy implies elevated genetic loading on the non-cognitive component.

The AR CAG repeat (androgen receptor, chrX) adds a functional data point. A hemizygous deletion at this locus estimates ~14 CAG repeats — approximately the 3rd percentile, meaning higher androgen receptor sensitivity than ~97% of the population. Shorter AR CAG repeats confer greater transcriptional activity of the androgen receptor — pleiotropic effects across multiple phenotypes through a single quantitative change in a transcription factor's activation domain.

Four Stories This Genome Tells

Bronze Age meets Mesolithic.  The paternal Y (R1b-DF27*) dates to the Bell Beaker Bronze Age expansion into Iberia ~2500 BCE. The maternal mtDNA (H3/H3g) dates to the post-glacial refugium re-expansion — more than 10,000 years ago. A 7,500-year asymmetry separates these lineages, and it is the canonical Iberian story: steppe males replaced prior Y-lineages during the Bronze Age while Mesolithic female continuity persisted across the same territory.

Norman Atlantic colonialism.  R1b-DF27* is found at 8–15% in Normandy. The pedigree documents Jean de Béthencourt (1362–1425), who conquered the Canary Islands in 1402–1405 under Castilian sponsorship — one of the earliest episodes of European Atlantic colonialism, ninety years before Columbus. Cadet branches dispersed into Iberia, the Azores, and eventually colonial New Granada. Colombia's ~38% R1b-DF27* frequency is a direct signature of the colonial male bottleneck. The genome sits at an intersection: Bell Beaker expansion 4,500 years ago, Norman Atlantic colonialism 600 years ago, and Iberian seeding of Latin America 500 years ago.

A Mestizo genome, predominantly Iberian.  The corrected analysis places the autosomal genome at ~55–65% European, ~20–30% Native American, ~10–20% African — consistent with an elite Caleño family pattern. The Native American component (~25%) represents the Nasa, Quimbaya, and Zenú peoples of the Cauca Valley. The African component (~12%) carries the history of enslaved people brought to the Cauca Valley's sugar economy from the 16th century onward. The pedigree traces 18 generations back to Toledo, Spain, before 1492 — the Hurtado family held civic positions in the Jewish community there up to the Alhambra Decree.

Non-cognitive genetic architecture.  The persistent EA > cognitive PGS divergence (∆ ≈ +1.0–1.4 SD across three studies) implies elevated genetic loading on the non-cognitive component of educational outcomes — conscientiousness, grit, persistence, socioeconomic pathway facilitation — relative to the g-factor component. The finding is coherent for someone who achieves formal credentials in philosophy and finance while operating outside conventional academic tracks and institutional structures.

Download the Genome — CC0

The full variant call file is released under CC0. No rights reserved, no restrictions. Use it for polygenic scoring, ancestry research, population genetics, or anything else.