Integrating Non-Contact 3D Point Cloud Phenotyping and Multi-Locus Genome-Wide Association Studies to Uncover the Genetic Architecture of Body Size in Pigs

Non-contact 3D point cloud phenotyping reduces the time and animal disturbance of manual livestock measurement, yet the genetic architecture of automatically derived body conformation traits remains underexplored. Here, we extracted body length (BL) and body height (BH) from 3D point cloud data of 510 Wens Black pigs and mapped their genetic basis. Pearson correlations between automated and manual measurements were 0.79 for BL and 0.74 for BH, with mean absolute percentage errors of 2.30% and 3.55%, respectively. The corresponding genetic correlation estimates were 1.00 ± 0.13 for BL and 0.92 ± 0.14 for BH. Combined single- and multi-locus genome-wide association study (GWAS) identified 43 candidate quantitative trait nucleotides (QTNs), including the BL-associated SNP detected at the suggestive significance level by the single-locus model and 42 additional loci at LOD ≥ 3. Among these, eight BL-associated and six BH-associated QTNs were detected by at least two multi-locus methods, including two BL-associated QTNs detected by all four methods. The maximum estimated proportions of phenotypic variance explained by individual QTNs were 6.00% for BL and 4.71% for BH. Functional enrichment implicated skeletal and cellular development pathways, prioritizing ACKR3, PLCB1 and COL11A1 for BL, and BDNF, PLEKHM1, ANKRD11 and RHOBTB2 for BH, as positional candidate genes. These findings demonstrate that 3D point cloud phenotyping captures heritable body-size variation in pigs and provide loci and candidate genes for validation and future genomic breeding.