Genome-Wide Discovery and Evaluation of Kazakh Ancestry-Informative Markers

Kazakh populations occupy an important position within the genetic landscape of Central Asia and Inner Eurasia, yet compact ancestry-informative marker panels evaluated for Kazakh-related ancestry remain limited. We analyzed genome-wide single-nucleotide polymorphism data from 624 Kazakh individuals together with data from external reference populations. After harmonization, the discovery dataset comprised 1607 individuals from 41 population groups and 80,766 shared autosomal variants. Candidate ancestry-informative markers were ranked using allele-frequency differentiation and informativeness for assignment, followed by linkage disequilibrium filtering. The final panel comprised 385 markers. Because marker availability differed across external datasets, transferability assessment in two independent Kazakh cohorts (n = 224 and n = 132) was performed using the 341-marker subset shared across the internal dataset and both external cohorts. Principal component analysis showed that both external Kazakh cohorts overlapped substantially with the internal Kazakh cluster. In nearest-centroid analysis, 94.6% of internal Kazakh individuals and 92.0% and 91.7% of individuals from the two external Kazakh cohorts, respectively, were assigned as Kazakh-related. Among non-Kazakh reference individuals used in the discovery framework, 4.1% were assigned as Kazakh-related; this internal, selection-informed estimate was mainly driven by genetically close Central Asian and Inner Eurasian populations. These findings support the transferability of the shared marker subset across independent Kazakh datasets, while also showing limited discrimination between Kazakhs and several closely related neighboring populations.