The lactase three-layer stack

The lactase locus interpreted across three layers, material γ, methylation, and 3D structure, leaves no silent gray zone. Every quantity is graded: read from structure, verified by measured data, established in the literature, dataset-named open, or a category boundary. Of twenty quantities, seventeen are positively evidenced.

DNA is read structurally, not linearly: the helix and folding bring distant positions together. At the lactase locus the material γ is conserved (|Δγ| = 0.034), a local CpG hotspot is the methylation substrate (the LCT promoter is unmethylated at β ≈ 0.004 in small-intestine epithelium where lactase is expressed, versus β ≈ 0.84 in non-expressing tissues), and the −13910 enhancer loops to the promoter to recruit demethylases. All twenty quantities are graded forced, verified, literature, open, or boundary, with none left silent.

DNA is read structurally, not linearly

The gray zone was a reading that treated each regulatory region as an isolated linear window and averaged it. But the helix and folding bring distant positions together, so they must be read together.

At the fine scale, a ~10.4 bp helical period sets nucleosome rotational positioning and so the accessibility of the methylation substrate; this arrangement signal is present in the LCT promoter at the 100th percentile against a composition-matched shuffle, and a single mean γ sees none of it. At the coarse scale, the −13910 enhancer sits ~14 kb from the LCT promoter on the linear sequence but is brought into contact by chromatin looping.

⚠ Superseded (v1.9). The reading just above — a ~10.4 bp helical period scored at the 100th percentile against a composition-matched shuffle — is an absolute, global periodicity statistic: a composition surrogate for nucleosome spacing read from an arbitrary window edge, not a statement about whether this element can contact a specific partner. The element-level structural read is the anchor-relative phase contact_competent of §13 (whether a motor and its nearest anchor sit on the same rotational helical face). The global periodicity is kept only as a descriptive composition fact; it is retired as a structural/contact claim — see the §8 retired register.

Three layers on one locus

The lactase switch is set by material, methylation, and 3D structure together. These three are reading channels, not a departure from the two-layer interpretation: each carries a Layer-1 form that structure fixes, while absolute magnitude stays the open Layer 2. Here the material γ is conserved (|Δγ| = 0.034), a CpG hotspot (O/E 0.88) is the methylation substrate, and the helical and looping structure gates accessibility and brings the enhancer into contact.

Adding methylation to the read-only γ gives the same material two opposite, discontinuous, hysteretic outcomes by genotype. Adding structure makes the LCT promoter and MCM6 enhancer one functional unit rather than two separate windows.

External measurements support the methylation layer

Methylation tracks the lactase switch in the predicted direction. In the human methylation atlas the LCT promoter is essentially unmethylated at β ≈ 0.004 in small-intestine epithelium, where lactase is expressed, and methylated at β ≈ 0.84 in non-expressing tissues such as lung, kidney, and breast.

ENCODE confirms the surrounding structure: the LCT promoter overlaps open-chromatin DNase clusters, and the −13910 region is annotated as a distal enhancer. The enhancer's bulk methylation is not tissue-specific, which refines the picture: the methylation switch is at the promoter, while the enhancer acts through the −13910 variant.

The literature confirms the model's structure

The established lactase mechanism matches the three-layer coupling this model posits. Age-dependent, genotype-dependent methylation gain at the LCT promoter is documented, with the −13910*T allele blocking it.

The −13910 enhancer physically contacts the LCT promoter via chromatin looping, and the bound transcription factors recruit demethylating enzymes to the promoter. That is exactly the order this model uses: structure controls methylation, which sets the switch.

An exhaustive ledger, and what no gray zone means

No gray zone means nothing is silent or vague, not that there are zero unknowns. Each of the twenty quantities of the lactase interpretation is graded: twelve read from structure, three verified in-package against measured data, two established by peer-reviewed measurement, one open but naming the dataset that would close it, and two category boundaries.

The one open item is a dataset-named refinement of an already-verified claim. The two boundary items are the framework's defining limits, the Layer-1 versus Layer-2 split and the DNA versus above-DNA split, which are the point of the framework rather than gaps in it.