Skip to main content
. 2011 Jan;35(1):19–45. doi: 10.1002/gepi.20547

Table VII.

Parameter estimates and log likelihoods from fitting various models to a single example data set

Maximum likelihood estimates of parameters representing effects due to

Child's genotype Mother's genotype Imprinting Maternal/fetal interactions




Row index Model fitted R1 R2 ρ1 ρ1 S1 S2 η1 η2 Im Ip Later Im Later Ip = jm γ11 γ22 γ01 = µ0 γ21 = µ2 µ jc Minus null loglik Minus max loglik Twice diff in logliks
1 EMIM imprinting (original Im) 3.86 24.43 1.16 1.50 0.35 4424.31 4186.53 475.56
2 EMIM imprinting (original Ip) 1.35 3.00 1.16 1.50 2.85 4424.31 4186.53 475.56
3 EMIM imprinting (later Weinberg [1999b] Im) 3.86 8.56 1.16 1.51 0.35 4424.31 4186.53 475.56
4 EMIM imprinting (later Weinberg [1999b] Ip) 1.35 8.56 1.16 1.51 2.86 4424.31 4186.53 475.56
5 Sinsheimer et al. [2003] 1A 1.24 5.44 2.00 2.30 4.89 4424.31 4176.68 495.26
6 EMIM 5df (single interaction γ11) 5.67 6.81 1.66 0.82 0.20 4424.31 4166.90 514.81
7 Palmer et al. [2006] 1.14 6.81 1.66 0.82 0.20 4424.31 4166.90 514.81
8 EMIM 5df (single interaction γ22) 3.08 14.07 0.55 0.82 0.42 4424.31 4197.28 454.07
9 Sinsheimer et al. [2003] 1B 1.03 5.91 2.02 1.27 6.20 3.48 4424.31 4165.14 518.35
10 EMIM (γ11 and γ22) 6.41 5.89 2.03 0.71 0.16 1.79 4424.31 4165.14 518.35
11 Parami et al. [2008] and Li et al. [2009] 1.50 5.90 1.11 1.55 2.16 1.99 4424.31 4161.08 526.45
12 Palmer et al. [2006] with Ip added 0.75 5.89 2.21 1.55 2.15 0.25 4424.31 4161.08 526.45
13 EMIM 6 df (Ip and γ11) 3.00 2.74 2.21 1.54 2.15 0.25 4424.31 4161.08 526.45
14 EMIM 6 df (Ip and γ22) 0.77 1.80 1.25 4.82 5.28 0.17 4424.31 4168.48 511.66
15 Sinsheimer [2003] 1A with Ip added 1.15 5.55 2.00 2.42 1.34 3.84 4424.31 4176.24 496.15
16 EMIM saturated (Ip, γ11 and γ22) 2.13 2.27 2.04 2.22 2.82 0.33 0.59 4424.31 4160.38 527.86
17 Sinsheimer [2003] 1B with Ipadded 0.69 6.41 2.04 1.31 2.86 3.04 5.25 4424.31 4160.38 527.86

EMIM, Estimation of Maternal, Imprinting and interaction effects using Multinomial modeling.