! MX EXAMPLE: Traditional Path Coefficients ACE/ADE Model ! ! BMI data from Australian Twins (Martin & Jardine, 1986) ! BMI estimated as 7 x log(wt/ht*ht), ! Where wt is weight in kg, ht is height in meters ! ! This univariate example uses only female twin data from young cohort #NGroups 3 Group 1: define model parameters and compute squared quantities Calculation_Group Begin Matrices; W LOwer 1 1 fixed !nonadditive/dominance genetic effects X LOwer 1 1 free !additive genetic effects Y LOwer 1 1 free !shared environmental effects Z LOwer 1 1 free !unique environmental effects H FUll 1 1 !fixed by default when not specified Q FUll 1 1 End Matrices: Matrix H !fixes value in matrix H to .5 .5 Matrix Q !fixes value in matrix Q to .25 .25 Begin Algebra; A = X*X' ; !A = a**2 = additive genetic variance component D = W*W' ; !D = d**2 = nonadditive genetic variance component C = Y*Y' ; !C = c**2 = shared env variance component E = Z*Z' ; !E = e**2 = nonshared env variance componet End Algebra; Label Row W dom_gen !labels for output Label Row X add_gen Label Row Y shr-env Label Row Z uni-env Label Row D dom_gen Label Row A add_gen Label Row C shr-env Label Row E uni-env Start .5 All !3 x .5squared = .75, roughly the observed Vp End Group 2: Young Female MZ Twins DAta_Group NInput=2 NObs=534 LAbels BMI_T1 BMI_T2 CMatrix SYmmetric !input/observed MZ cov matrix .7247 .5891 .7915 Matrices = Group 1 Covarinace A+C+D+E | A+C+D _ !expected MZ cov matrix based on model A+C+D | A+C+D+E / Options RSiduals ND=3 !options for output End Group 3: Young Female DZ Twins DAta_Group NInput=2 NObs=328 LAbels BMI_T1 BMI_T2 CMatrix SYmmetric !observed DZ cov matrix .7786 .2461 .8365 Matrices = Group 1 COvarinace A+C+D+E | H@A+C+Q@D _ !expected DZ cov matrix H@A+C+Q@D | A+C+D+E / Options RSiduals ND=3 End