Disentangling mature cow weight and body condition score: a comparative single-step GWAS of phenotypic pre-adjustment...

Mature cow weight (MWT) is a trait genetically correlated with body condition score (BCS). Previous research has shown that sire rankings can shift depending on how BCS is accounted for, indicating that different modeling strategies can influence selection outcomes. The recursive modeling approach has been established as a method for obtaining MWT that is genetically independent of BCS, providing an alternative to phenotypic pre-adjustment. The objective of this study was to determine whether different modeling approaches capture different genetic architectures or merely produce statistical artifacts. Genome-wide association studies (GWAS) and functional genomic analyses were performed to compare the genomic architecture of phenotypically pre-adjusted MWT (MWTadj) with MWT that is genetically independent of BCS, obtained using the recursive approach (MWTRM). A total of 42 significant SNP across 8 chromosomes were identified for MWTadj and 44 SNP across nine chromosomes for MWTRM, with 28 SNP shared between models. These variants corresponded to 107 annotated genes in MWTadj and 137 in MWTRM, including 62 shared genes. Major association signals were concentrated on BTA20, BTA7, and BTA14 for both models, with all significant SNP jointly explaining 3.93% of the total additive genetic variance for MWTadj and 4.29% for MWTRM. The Pearson correlation coefficient of estimated SNP effects between the models was 0.76, while the correlation of genomic estimated breeding values was 0.87. Compared to MWTadj, which shared 19 genes with unadjusted MWT, the MWTRM shared 33 genes. In addition, 28 of the 31 pathways identified for MWTRM were also identified for unadjusted MWT, whereas no Gene Ontology pathways were shared between MWTadj and unadjusted MWT. The MWTRM was associated with genes annotated to growth, skeletal development, feed efficiency, and carcass-related traits, whereas MWTadj identified a distinct set of genes and pathways. Despite these differences, MWTadj and MWTRM converged on similar core biological signals, highlighting that they effectively capture the primary genetic drivers of MWT independent of BCS.