Summary of studies comparing estimates of quantitative genetics parameters using pedigree-free methods.

1Method: 1 = Ritland’s method; 2 = maximum likelihood on relatedness classes; 3 = sibship or pedigree reconstruction; 4 = animal model with marker-based relatedness matrix; 5 = genomic selection.2SSR = microsatellite; SNP = single nucleotide substitution.3Method used to estimate pairwise kinship coefficients using molecular markers: LR = Lynch & Ritland [16]; QG = Queller & Goodnight [15]; W = Wang [17]; R = Ritland [21]; SA = % of shared alleles [118]; GRM = genomic relationship matrix [119]; IBS(i,j) = probability of identity by state [120]. For method 3, this column indicates the method of pedigree reconstruction: MCMC (maximizes the pairwise likelihood ratios of being full siblings or unrelated [121]); CERVUS [122]; COLONY [123]; PAPA [124].4The bias is defined as the difference between heritability estimated from the pedigree (ΦA) and heritability estimated using molecular markers in one of the five methods presented marker–ped. When no information was available on ped estimated using pedigrees, the bias was measured relative to estimated using method 3 (pedigree reconstruction), as indicated in the column “Note”. When several traits were analyzed (column “traits”), we present the average bias.5Mating regime: “Out” stands for outcrossing, “Mix” for mixed mating, “Self” for predominantly selfing and “C” for clonality.6Single Locus Probe.