We introduce MXG+, a structural grammar framework that quantifies organizational properties of genomic sequences independently of sequence content, evolutionary models, or functional annotation. Applied to five archaic hominin genomes (Neanderthal, Denisovan) and 49 modern human individuals spanning the full range of known Neanderthal introgression (0–4.5%), MXG+ reveals that genomic variant sequences carry a measurable structural grammar signal that decomposes into three separable biological dimensions: (1) population bottleneck history (42.6% of explained variance); (2) archaic introgression, which contributes a statistically significant secondary signal surviving heterozygosity removal (Spearman ρ=+0.52, p=0.0001); and (3) deep ancestral divergence (15.4% of explained variance), which explains the anomalous structural position of Southern African Khoisan populations. Adding MXG+ scores to heterozygosity improves introgression prediction by ΔR²=+0.163 (F=9.74, p=0.003). A blind test on 20 populations never used in model construction confirms generalization (Spearman ρ=+0.56, p=0.010). Structural scores are consistent across three independent genomic representations (variant, coding, methylation) and the methylation F-component (transition clustering) increases monotonically from great ape (0.375) to archaic hominin (0.556) to modern human (0.578). Cross-referencing with published engraving structural analyses (Molina 2026a, 2026b), the same F-component discriminates chimp from hominin in both behavioral output and molecular substrate across 380,000 years of hominin evolution. Analytical methodology withheld pending patent filing (FMX-IP-2026); full methodology deposited separately under restricted access. All statistical results fully reproducible from publicly available genotype data and reported scores. Priority date: April 5, 2026.
Juan Gabriel Molina (Sun,) studied this question.