This study provides a comprehensive characterization of the whole genomes of the Kazakh breeds of the Jabe and Аdai, types, as well as the Mugalzhar breed. We performed whole genome sequencing on the BGI platform, mapping the obtained data to the EquCab3.0 horse reference genome. The sequencing results revealed a large number of genetic variants in the genomes of the studied horses, indicating a high level of genetic polymorphism. The largest number of variants was found in intronic regions, as well as in untranslated regions. Among the identified variants, single nucleotide polymorphisms were found to be prevalent both before and after filtering, as well as insertions/deletions, which occur much less frequently. Special attention was paid to the SWT1, PPFIA1, SHROOM2, ANOS1, OR4D10, USP25, and ZNF74 genes, which contained more than 100 variants each. It is assumed that these genes can influence metabolism, endurance, reproductive function, sense of smell, immune response and regulation of gene expression in horses. The obtained data represent a valuable resource for future studies on the adaptive potential of Kazakh horse breeds, as well as for the development of conservation strategies and the optimization of breeding programs.

Kazakh horses represent an important cultural heritage of the Kazakh people, shaped over thousands of years across the territory of present-day Kazakhstan. Long-term adaptation to extreme natural conditions has led to the development of a unique set of characteristics that ensure the high survivability, ecological plasticity and economic value of these horses. Long- and short-read sequencing was performed for three breeds (Adai, Zhabe, Mugalzhar), while Naiman was sequenced exclusively with nanopore long-read data. The resulting reads were assembled, and the genomes were subsequently annotated. The assembled genomes showed high contiguity with an average contig N50 value of 9.68 Mb (9,681,711 bp) for Adai, Zhabe and Mugalzhar, while the resulting contig N50 was 17.09 Mb (17,090,216 bp) for Naiman, and average assembly length was 2.42 Gb (2,442,471,376 bp). These are individual reference assemblies that establish foundations for future comparative genomics and pangenome development. Gene annotation yielded 19,584, 19,941, 18,695 and 20,572 protein-coding genes for Adai, Zhabe, Mugalzhar and Naiman, respectively. These assemblies complement existing assemblies for future genetic analysis, research into the geographic and breed origins of domestic horses, and breed conservation efforts, and will contribute to ongoing equine pangenome efforts.