Abstract: Chinese indigenous pig breeds, despite serving as a crucial reservoir of genetic variation for adaptation, product quality, fertility, and disease resistance, are overlooked in global swine breeding efforts. Despite the capability of large-scale sequencing to facilitate high-resolution genetic characterization, problems remain in the practical application of genomic findings. Genomic selection (GS) improves complex traits but is hindered in local populations by restricted effective population sizes, inadequate phenotyping, inbreeding risks, and varied breeding goals. Recent advancements tackle these challenges through three approaches: the establishment of multi-trait genomic selection indices utilizing genomic optimal contribution selection (GOCS) and genomic mating (GM) to concurrently enhance genetic gain and diversity, the augmentation of reference populations via cost-efficient genotyping and imputation, and the incorporation of multi-omics and machine learning to enhance predictive accuracy in conditions of data scarcity. Future advancements will depend on the incorporation of multi-source data inside GS frameworks, encompassing high-resolution phenotyping and artificial intelligence (AI), to improve cross-population prediction and trait analysis. By summarizing current advances in genomic selection for Chinese indigenous pig breeds, this review outlines opportunities and challenges for integrating genomic tools into genetic resource conservation and breeding programs.

Key words: Chinese local pig, genomic selection, adaptation, genetic diversity, selection index, multi-omics