The evidence ladder begins with a reporter assay: 561,410 human–ape substitutions were tested and 15,077 loci showed significantly different regulatory activity. Hybrid human–ape cells then revealed 4,463 human-specific differences in nearby, or cis, expression. Pathway analysis found that 17 of 21 differently expressed glycosaminoglycan genes were downregulated, and human joint tissue contained approximately one-quarter to one-third as much glycosaminoglycan as joint tissue from non-human great apes. Together these observations support regulatory remodeling of cartilage biology, but they do not prove that any one substitution changed skeletal shape, caused bipedalism or produced arthritis.

1
SequenceReporter assay
Can a human DNA letter change regulatory activity relative to its ancestral great-ape version?
561,410 substitutions tested; 15,077 loci showed significantly different activity.Does not establish: a change in bone shape, walking or disease.
2
Shared cellsHybrid-cell comparison
Does the difference persist when human and ape genomes operate in the same cellular environment?
4,463 human-specific differences in cis-regulated expression were reported.Does not establish: how a complete, moving joint develops or functions.
3
PathwayConverging signals
Do many regulatory and expression differences point toward the same biological system?
Glycosaminoglycan biology emerged repeatedly; 17 of 21 differently expressed GAG genes were downregulated.Does not establish: which skeletal trait selection favored.
4
TissueJoint measurement
Does actual cartilage show a corresponding biochemical difference?
Human joints contained approximately one-quarter to one-third as much GAG as joints from non-human great apes.Does not establish: that one variant caused the difference or that reduced GAG caused arthritis.

Read upward carefully: Convergence makes the broader hypothesis more plausible; it does not turn association into single-variant causation.