Age-based partitioning of individual genomic inbreeding levels in Belgian Blue cattle

Table 2 Comparison of models used to estimate genomic inbreeding coefficients with different numbers of HBD classes (from 1 to 4)

Panel density	N^a	Number of fitted HBD classes	Mean F_G	Correlation with F ^b_G-8192	Median of estimated R _k rates per HBD class
Panel density	N^a	Number of fitted HBD classes	Mean F_G	Correlation with F ^b_G-8192	1st HBD class	2nd HBD class	3rd HBD class	4th HBD class
LD	634	1	0.058	0.982	15
LD	0	2	0.061	0.998	11	106
LD	0	3	0.061	0.999	11	104	162
LD	0	4	0.061	0.999	10	42	150	175
50 K	289	1	0.083	0.983	41
50 K	345	2	0.094	0.999	15	198
50 K	0	3	0.094	1.000	14	173	238
50 K	0	4	0.094	1.000	11	64	240	243
HD	0	1	0.297	0.999	1214
HD	0	2	0.302	1.000	60	1679
HD	629	3	0.303	1.000	22	392	1887
HD	5	4	0.303	1.000	19	342	1823	1914
WGS	0	1	0.354	1.000	3740
WGS	0	2	0.359	1.000	577	8158
WGS	27	3	0.359	1.000	55	1009	8192
WGS	23	4	0.359	1.000	21	206	1104	8192

The R _k rates of each HBD class were estimated for each individual and for each SNP density (LD, 50 K, HD panels or whole-genome sequence data). The table reports which models are selected as best based on the BIC criterion, the average F_G and its correlation with a reference F_G obtained with a model using 13 HBD classes
^aN = number of individuals with the corresponding model selected as best based on the BIC
^bThe reference inbreeding coefficient F_G-8192 is obtained with a Mix14R model and the same SNP density

ISSN: 1297-9686