Human metaphase chromosomes follow a tight length-to-DNA scaling rule
A new quantitative framework for human chromosome packing links metaphase length to DNA content with unusual precision. Across 24 human chromosomes in GRCh38, the relationship between mean metaphase length and genomic DNA content was nearly linear, with an OLS fit reaching an R-squared of 0.998. In practical terms, larger chromosomes are not merely longer in a casual sense; they scale in a way that is mathematically regular enough to define a compaction density of 33.4 nm of metaphase per megabase of DNA, with a coefficient of variation of just 2.9 percent.
This is a physical measurement of how the genome is folded into the mitotic state. The result matters because metaphase chromosomes have long been treated as visually discrete cytogenetic objects, yet the data show a continuous size law underlying those objects. Metacentric, submetacentric, acrocentric, and sex chromosomes all fall within the same overall scaling logic, even though they do not share identical architecture.
One chromosome did not simply sit as an exception at the margin. It fit into the same rule set, but with residuals that reveal where chromosome-specific structure still shapes the final mitotic form.
Heterochromatin changes the residuals, not the overall scaling logic
Residual analysis exposed the chromosomes that deviate most from the simple length-to-DNA line. Under the GRCh38 reference, chromosomes 9, 16, and Y stood out after the transition to T2T-CHM13 v2.0, where the largest heterochromatin additions produced large new negative residuals. In plain terms, when previously unresolved repetitive regions were added to the genome assembly, the apparent mismatch between DNA mass and metaphase length shifted in a predictable direction.
The p-arm also carried important information. Length residuals correlated with p-arm Mb content, with Pearson r = 0.65 and p = 5.6e-4. Acrocentric chromosomes 13 to 15 clustered at low p-arm Mb and negative residuals, a pattern consistent with the idea that arm composition influences how much metaphase space a chromosome occupies for a given amount of DNA. This is not a mere cartographic detail. Heterochromatin density, repetitive sequence content, and arm asymmetry all alter the way chromatin can be compacted into the mitotic scaffold.
For bedside genetics, that logic is familiar even if the setting is different. Structural chromosome behavior often reflects sequence context, not just raw length. The same principle appears here at the scale of whole metaphase chromosomes.
The data favor a 7.6 Mb unit over the older late-metaphase estimate
A central test in the analysis compared observed metaphase lengths with the helical-turn count implied by a previously proposed late-metaphase parameter of 12 Mb per turn. The human data instead supported an OLS slope of 0.395 μm per turn and implied a DNA content of 7.6 Mb per turn, described as 1 MCU. That estimate is materially smaller than the earlier 0.25 μm per turn benchmark, and it shifts the assumed geometry of how chromatin is organized during late metaphase.
The biological implication is straightforward. If each higher-order turn contains less DNA than previously assumed, then the chromosome is built from smaller, more numerous physical modules than older models predicted. That changes how investigators think about the relationship between loop arrays, scaffold attachment, and the visible metaphase axis.
A second analysis linked metaphase length to real Hi-C boundary counts. Using Rao 2014 GM12878 HiCCUPS loops and Arrowhead contact domains, per-chromosome Hi-C boundary regression produced slopes of 8.8 and 8.9 nm per boundary, with R-squared values of 0.84 and 0.78. The signal is strong enough to suggest that chromatin contact architecture and metaphase packing are not separate stories. They are coupled descriptions of the same folding problem at different scales.
Chromosome pairs scale with extraordinary regularity
Pairwise inter-chromosomal scaling strengthened the framework further. Across 276 chromosome pairs, the slope reached 32.75 nm/Mb with an R-squared of 0.996. The fit is so tight that the pairwise relation reads less like a cloud of points and more like a governing physical law. Even the original chr 1 versus chr 2 observation, represented by 6.76 Mb and 0.232 μm, sits inside the same framework rather than outside it.
The quantization test is equally revealing. When putative domain count was estimated by dividing length by 0.23 μm, the values did not cluster at integers. In other words, there was no detectable quantization at the 0.23 μm scale. This argues against a simple model in which metaphase chromosomes are built from rigid, integer-numbered physical units at that size. The chromosome axis behaves as a continuum governed by scaling, not as a staircase of identical length blocks.
The MCU framework closes a missing gap in chromatin hierarchy
The chromatin hierarchy presented in the figure set runs from B-form DNA at L1 to the whole metaphase chromatid at L10. Within that sequence, the MCU occupies L9 and bridges a quantitative gap between mitotic loops and whole chromosomes. The value of the framework is not only that it introduces a new number. It supplies a missing conversion factor between cytogenetic appearance, polymer physics, Hi-C organization, and evolutionary scale.
The companion “33 nm/Mb dictionary” turns the observation into a practical reference structure. By translating between DNA content, metaphase length, and higher-order chromatin modules, the framework gives researchers a common language for discussing compaction across experimental systems. An interactive web-based MCU calculator is also available online and archived, extending the analysis beyond the static figures.
In a field where chromosome packing has often been described qualitatively, this study replaces vague language with measurable relationships. The strongest limitation is also its clearest boundary: the framework is built from inference across reference assemblies, cytogenetic length estimates, and contact-domain mappings, not direct molecular visualization of every packing transition in living human cells. The next tests will have to determine how robust the 7.6 Mb unit remains across cell states, species, and perturbations of mitotic architecture.