Ashkenazi Population Structure Reassessed: Reference-Panel Sensitivity and the Europe–Levant Cline

This revised manuscript reassesses Ashkenazi population structure by testing how reference-population design affects inference of a Europe–Levant genetic cline. It combines author-generated qpAdm and autosomal FST analyses with a linked Global25 sensitivity suite comprising fifteen experiments, including matched omission tests, reciprocal Italian Jewish models, regional Ashkenazi targets, and ancient Levant positive controls.

The central result is that omitting Italian Jewish, Southern Italian, Sicilian, Maltese, and Aegean comparators worsens model fit and redistributes their signal into north-shifted Italian and Levantine proxies. In matched Experiments 10–12, removing the wider central Mediterranean source space worsens fit by 47.18%, raises the Lombardy coefficient from 9.8 to 38.0, and raises the Levantine-continuity coefficient from 22.8 to 52.0. The results show that an apparent Europe–Levant cline can be created or exaggerated by comparator omission.

The manuscript interprets Global25 coefficients as panel-dependent similarity weights rather than literal ancestry proportions. The complete reproducibility package is available on GitHub at https://github.com/StevenParker-collab/ashkenazi-dna-revised-pci-reproducibility and as a frozen Zenodo archive at https://doi.org/10.5281/zenodo.21383280.