Exploring the genetic architecture and improving genomic prediction accuracy for mastitis and milk production traits in dairy cattle by mapping variants to hepatic transcriptomic regions responsive to intra-mammary infection

Table 2 Top five predictive genomic features for mastitis, protein, milk and fat yield in Holstein cattle

Trait	Time (h)^a	\({\text{FDR}}_{\exp }^{\text{b}}\)	Log₂(FC)^c	\({\text{P}}_{\text{set - test}}^{\text{d}}\)	SNP_f (%)^e	\({\text{H}}_{\text{f}}^{2}\) (%)^f	\(r_{\text{GFBLUP}}^{\text{g}}\)	bias ^h	Δr ⁱ
Mastitis	9	5 × 10⁻²	NA^j	0.013	6.36	25.60	0.520	0.872	0.016
	9	5 × 10⁻²	>1	0.027	2.32	13.71	0.519	0.872	0.015
	6	5 × 10⁻²	NA	0.040	5.92	19.81	0.519	0.873	0.015
	6	10⁻²	NA	0.043	4.68	18.83	0.518	0.871	0.014
	6	10⁻³	NA	0.034	3.54	15.39	0.518	0.871	0.014
Protein	48	10⁻⁶	>2	0.021	<0.01	1.85	0.622	0.783	0.020
	48	10⁻⁸	>2	0.029	<0.01	1.75	0.621	0.782	0.019
	48	10⁻²	>2	0.023	0.02	3.28	0.621	0.779	0.019
	48	10⁻⁸	>1	0.027	<.01	1.71	0.621	0.782	0.019
	48	10⁻¹⁰	>2	0.026	<0.01	1.37	0.620	0.782	0.018
Milk	6	10⁻²	NA	0.026	4.68	31.90	0.651	0.863	0.016
	6	10⁻³	NA	0.027	3.54	26.82	0.651	0.865	0.016
	6	10⁻³	<−1	0.024	1.76	19.74	0.650	0.862	0.015
	6	10⁻⁶	<−2	0.022	0.28	12.49	0.649	0.866	0.014
	6	10⁻²	<−1	0.030	2.49	25.39	0.649	0.859	0.014
Fat	6	10⁻⁶	<−2	0.027	0.28	16.28	0.629	0.804	0.022
	6	10⁻³	<−2	0.028	0.33	17.76	0.626	0.800	0.019
	6	10⁻²	<−2	0.032	0.36	18.57	0.625	0.798	0.018
	6	5 × 10⁻²	<−2	0.032	0.37	18.51	0.625	0.799	0.018
	9	10⁻⁶	>1	0.055	0.84	20.94	0.621	0.815	0.014

^aTime points post intra-mammary infection with E. coli LPS
^bFDR values used to define genomic features from RNA-Seq analysis
^cLog₂(fold-change) values used to define up- (down-) regulated genomic features from RNA-Seq analysis
^dP values from SNP set test on HOL training population
^eProportion of SNPs in genomic features over the whole genome
^fProportion of the total genomic variance explained by genomic features
^gPrediction accuracy with GFBLUP
^hThe regression coefficient of de-regressed proofs (DRP) on predicted genomic breeding values (GEBV)
ⁱThe change of prediction accuracy with GFBLUP relative to GBLUP
^jThe genomic feature defined without log₂(fold-change)

ISSN: 1297-9686