MTDFNRM MTDFPREP pedigree file data file - PowerPoint Presentation

1. MTDFNRM MTDFPREPpedigree file data filefree format free format MTDF11MTDF44MTDF13MTDF56MTDFRUNMTDF21MTDF22MTDF50MTDF51MTDF52MTDF66MTDF4MTDF54MTDF58MTDF59MTDF68MTDF72MTDF76MTDF77MTDF78MTDF79MTDF67CONTINUERESTART••••••

2. MTDFREML Output Files (* = optional)MTDFNRM.EXE MTDF11 No. animals, Recoded and original animal ID * MTDF13 No. animals, Recoded and original ID and F for animal, sire, dam MTDF44 BINARY, elements of A-1 MTDF56 Summary for MTDFNRM MTDFPREP.EXE * MTDF21 Levels of fixed factors and description * MTDF22 Levels of other random factors and description MTDF50 Markers for setting up MME MTDF51 Binary Basically recoded data file MTDF52 Binary for setting up MME MTDF66 Summary for MTDFPREP

3. MTDFRUN.EXE MTDF4 Updated start file MTDF54 Current simplex, -2 logL and VC MTDF58 Reorder file for SPARSPAK MTDF59 File of user imposed constraints MTDF68 History of -2 logL and VC * MTDF72 F, PBV, SEP and accuracy by animal MTDF76 Summary for MTDFRUN, VC and genetic parameters * MTDF77 Fixed effect solutions * MTDF78 PBV solutions * MTDF79 Other random effect solutions * MTDF67 Blocks of inverse, F-tests Contrasts, t-tests Expected values of solutions

4. COMPILING and LINKING FORTRAN code needs to be converted to an executable program in 'machine' language Portable Flexible (match resources to problem) Need a FORTRAN compiler MS Powerstation >>Digital>> Compaq Lahey G77(DOS, LINUX) Usual: *.FOR >> *.OBJ (subroutines, programs) *.FOR >> *.OBJ >>LINK>> *.EXE (subroutines linked to program)

5. Optimization switches (debugging vs speed) Switches in SETUP.BAT and COMPILE.BAT are for speed not debugging "Include" statement for a file The file is compiled with rest of code PARAM.DAT sets DIMENSIONS of analysis (matrices, etc.) After changes in PARAM.DAT must recompile

6. MAIN PROGRAMS with SUBROUTINES MTDFNRM.FOR Pedigree file to A-inverse MSTIME.FOR (Timer subroutines) MTDFSUB.FOR (Many subroutines) MTDFPREP.FOR Converts data to standard form (equation no., etc.) Include 'PARAM.DAT' MSTIME.FOR MTDFSUB.FOR

7. MTDFRUN.FOR Sets up MME, solves, SIMPLEX algorithm for VC Include 'PARAM.DAT' MSTIME.FOR MTDFSUB.FOR SPARS4.FOR (SPARSPAK subroutines: SPARS4A.FOR reorder, Choleski factor and solve) MTDFLIK.FOR (subroutine to calculate -2logL, etc.) Include 'PARAM.DAT' SETUP.BAT Most subroutines to *.OBJ COMPILE.BAT MTDFLIK.FOR to *.OBJ; MTDFNRM.FOR, MTDFPREP.FOR, MTDFRUN.FOR, compiled and linked to MTDFNRM.EXE, MTDFPREP.EXE, MTDFRUN.EXE

8. c This file is 'param.dat' with code that is included in compiling c mtdfprep.for, mtdflik.for, and mtdfrun.for PARAMETER( c no. traits * MAXTRT=2, c no. of integer variables on record (fixed and random effects), * MAXINTR=30, c no. of real variables on record (covariates and traits) * MAXR8=30, c no. of covariates for each trait * MAXCOV=1, c no. of regression coefficients for each covariate * MAXNFR=1, c no. of fixed effects for each trait * MAXFIX=3, c no. of levels for each fixed effect * MAXNFL=50,c no. of constraints for equations * MAXCONS=1, c no. of animals for each trait * MAXANIM=600,c maximum observations – needed for AI matrix * MAXOBS = 2000,c no. of additional uncorrelated random effects for each trait * MAXRAN=3, c no. of levels for each additional uncorrelated random effect * MAXNRL=500, c no. of non-zero elements in coefficient matrix used for c fspak version only + MAXNZE=525000,c length of hash vector used for fspak version only + NHASH=650000,c for sampling variances + MAXINV=40,c length of S vector for SPARSPAK version only * MAXORDS=1250000,c ------------------------ changes to here ----------------------------

9. rem Reminder that this file is 'setup.bat'; 'rem' is not active commandrem Note that '/G4' will be ignored FL32 /c /G4 /Op /D "NDEBUG" spars4a.forFL32 /c /W0 /G4 /Ox /Op /D "NDEBUG" spars4.forFL32 /c /G4 /Ox /Op /D "NDEBUG" mstime.forFL32 /c /G4 /Ox /Op /D "NDEBUG" mtdfsub.forcall compile.bat

10. rem Reminder that this is file 'compile.bat', 'rem' is not activeFL32 /G4 /Ox /Op /D "NDEBUG" mtdfnrm.for mtdfsub.obj mstime.objFL32 /G4 /Ox /Op /D "NDEBUG" mtdfprep.for mtdfsub.obj mstime.objFL32 /c /G4 /Ox /Op /D "NDEBUG" mtdflik.forFL32 /c /G4 /Ox /Op /D "NDEBUG" mtdfrun.forFL32 /G4 /Ox /Op /D "NDEBUG" mtdfrun.obj mtdflik.obj mstime.obj mtdfsub.obj spars4.obj spars4a.objrem '/G4' will be ignored

11. rem The is MOUSE.BAT a batch file that calls answer files for rem a complete run with one restart of MTDFREMLrem rem is a memo for DOSrem whatever is to the right of < is used as an answer filerem in place of answers typed on keyboardmtdfnrm<mousenrm.in rem copy to save mtdf56 to mtdf56.1 (on next run old mtdf56 is replaced)copy mtdf56 mtdf56.1 rem < answer file for mtdfprep--can be any namemtdfprep<mouseprp.in rem copy to save mtdf66 to mtdf66.1 which is log file for mtdfprepcopy mtdf66 mtdf66.1

12. rem < answer file for initial start for mtdfrun--can be any namemtdfrun<mouserun.in rem copy mtdf4 to mtdf4.1 for restart of mtdfrun with mtdf4.1 as answer copy mtdf4 mtdf4.1 rem copy to save mtdf76 to mtdf76.1 to check -2log L on next restartcopy mtdf76 mtdf76.1 rem < answer file from previous start updated with new estimatesmtdfrun<mtdf4.1 copy mtdf4 mtdf4.2rem copy to save to check -2log L and estimates for global converge copy mtdf76 mtdf76.2 rem < answer file to use previous estimates to solve MME andrem to calculate SE of contrasts and expectations of solutionsmtdfrun<mouserun.se rem copy mtdf67 to mtdf67.1 to save file of contrasts and expectations copy mtdf67 mtdf67.1 rem copy mtdf77 to mtdf77.1 to save fixed effects solutions copy mtdf77 mtdf77.1

13. 20101 11012 10101 0 This is 'mouse.ped' 11012 615 701 0 Animal Sire Dam Zero (not needed) 10101 0 0 0 20102 11012 10101 0 Does not need to be ordered 20103 11012 10101 0 20104 11012 10101 0 Can have duplicates 20112 11012 10101 0 20113 11012 10101 0 An ID > Sire and Dam ID 20114 11012 10101 0 20115 11012 10101 0 Zero ID is unknown 20302 10614 10506 0 10614 2315 2201 0 For sire model would have ped file 10506 1813 1701 0 20303 10614 10506 0 SireID 0 0 20304 10614 10506 0 20306 10614 10506 0 Can use field 2, and 4 twice 20312 10614 10506 0 20313 10614 10506 0 20314 10614 10506 0 20401 10813 10701 0 . . . . . . . . . . . . . .

14. 41601 31514 31602 0 41602 31514 31602 0 41603 31514 31602 0 41604 31514 31602 0 41612 31514 31602 0 41613 31514 31602 0 41614 31514 31602 0 41615 31514 31602 0

15. 0 0 for animal model,1 for sire/MGS model [This is 'mousenrm.in']41615 maximum id number represented in pedigree file, can be larger1 minimum id number represented in pedigree file, can be 0mouse.ped name of pedigree file1 write inbreeding file to mtdf13, 0 would not write that file4 no. integer fields in ped file (must be no reals before them)*1 position of animal (or sire for sire, sire of sire, mgs of sire) 2 position of sire (or sire of sire)3 position of dam (or MGS of sire)0 number of genetic (Westell) groups

16. Started 07:05:12.17 on 08/10/1998 FILE is MTDF56 from MTDFNRM+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ PROGRAM "MTDFNRM" - Calculate A-1 for "MTFRUN" and recode IDs for "MTDFPREP" Version to use Westell grouping strategy+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ OPTION FOR CALCULATION OF A-1 FOR ANIMAL SIRE DAM TYPE .... 0 FOR ANIMAL SIRE MGS TYPE ... 1 OPTION CHOSEN FOR THIS ANALYSIS = 0 MAXIMUM ID = 41615 MINIMUM ID = 1 PEDIGREE FILE OPENED, IUN33 = mouse.ped ************ REORDERED ANIMAL FILE OPENED, IUN11 = MTDF11 FILE FOR A-1 ELEMENTS OPENED, IUN44 = MTDF44 FILE FOR IDS AND INBREEDING COEFFICIENTS OPENED THIS FILE WILL CONTAIN ANIMAL, SIRE, AND DAM RECODED AND ORIGINAL IDS FOLLOWED BY THE INBREEDING COEFFICIENT FOR EACH NO. INTEGER FIELDS PER RECORD IN IUN33 = 4 ANIMAL ID IN POSITION ...... 1 SIRE ID IN POSITION ........ 2 DAM (MGS) ID IN POSITION ... 3 NO. OF GENETIC GROUPS FOR CALCULATION OF W = 0 The current time is: 07:05:12.72 NO. OF PEDIGREES READ = 309 NO. OF DIFFERENT ANIMALS = 329 INCLUDES NO. OF GENETIC GROUPS = 0

17. END OF FIRST PASS The current time is: 07:05:12.89 END OF SORT The current time is: 07:05:12.89 FIRST 10 REORDERED IDs 1 215 FIRST 10 REORDERED IDs 2 403 FIRST 10 REORDERED IDs 3 615 FIRST 10 REORDERED IDs 4 701 ID VECTOR WRITTEN IN ORDER TO IUN11 The current time is: 07:05:13.05 SIRE AND DAM IN PEDIGREE REORDERED IN IVECS AND IVECD The current time is: 07:05:13.33 CALCULATION OF A-1 FROM ANIMAL SIRE DAM (IOPT = 0) NON-ZERO HS ELEMENTS FOR NRM INVERSE = 1241 LOG DETERMINANT OF NRM = -210.71674289 NUMBER OF INBRED ANIMALS = 0... WITH AVERAGE INBREEDING COEF = .00000000 TOTAL NO. OF ANIMALS INCLUDING BASE AND GENETIC GROUPS = 329 ******* The current time is: 07:05:13.99 The elapsed time was: 00:00:01.27

18. 329 (No.) [ MTDF13 optional for an, sire, dam codes and F] RECODED ORIGINAL F = inbreeding AN SIRE DAM AN SIRE DAM AN SIRE DAM 1 0 0 215 0 0 .000 .000 .000 2 0 0 403 0 0 .000 .000 .000 3 0 0 615 0 0 .000 .000 .000 4 0 0 701 0 0 .000 .000 .000 5 0 0 814 0 0 .000 .000 .000 6 0 0 904 0 0 .000 .000 .000 7 0 0 1314 0 0 .000 .000 .000 8 0 0 1602 0 0 .000 .000 .000 9 0 0 1701 0 0 .000 .000 .000 10 0 0 1813 0 0 .000 .000 .000 11 0 0 2201 0 0 .000 .000 .000 12 0 0 2315 0 0 .000 .000 .000 13 0 0 2401 0 0 .000 .000 .000

19. 23 0 0 10404 0 0 .000 .000 .000 24 10 9 10506 1813 1701 .000 .000 .000 25 10 9 10512 1813 1701 .000 .000 .000 26 12 11 10602 2315 2201 .000 .000 .000 27 12 11 10614 2315 2201 .000 .000 .000 28 16 13 10701 2512 2401 .000 .000 .000 29 16 13 10715 2512 2401 .000 .000 .000 30 20 17 10801 3015 2904 .000 .000 .000 31 20 17 10813 3015 2904 .000 .000 .000 32 1 2 10902 215 403 .000 .000 .000 33 1 2 10917 215 403 .000 .000 .000 34 3 4 11002 615 701 .000 .000 .000 40 0 0 11316 0 0 .000 .000 .000 41 0 0 11404 0 0 .000 .000 .000 42 14 15 11505 2416 2503 .000 .000 .000 . . . 329 214 216 41615 31514 31602 .000 .000 .000

20. This is 'mouse.dat' Animal Sire Dam Gen Sex LS Lit LS* T-1 T-2_____________________________________________________ 20101 11012 10101 1 1 4 1 4.0 22.5 59.1 20102 11012 10101 1 1 4 1 4.0 22.6 52.4 20103 11012 10101 1 1 4 1 4.0 22.9 61.1 20104 11012 10101 1 1 4 1 4.0 23.0 57.9 20112 11012 10101 1 2 4 1 4.0 24.6 69.3 20113 11012 10101 1 2 4 1 4.0 26.4 66.4 20114 11012 10101 1 2 4 1 4.0 24.1 61.6 20115 11012 10101 1 2 4 1 4.0 24.3 68.3 20302 10614 10506 1 1 7 2 7.0 23.6 60.1 20303 10614 10506 1 1 7 2 7.0 24.2 60.8 20304 10614 10506 1 1 7 2 7.0 22.5 62.8 20306 10614 10506 1 1 7 2 7.0 22.3 59.4 20312 10614 10506 1 2 7 2 7.0 23.7 66.9 20313 10614 10506 1 2 7 2 7.0 27.5 72.9 20314 10614 10506 1 2 7 2 7.0 17.3 58.4 20401 10813 10701 1 1 2 3 2.0 20.3 58.1

21. 20402 10813 10701 1 1 2 3 2.0 21.3 58.7 20403 10813 10701 1 1 2 3 2.0 21.9 53.2 20404 10813 10701 1 1 2 3 2.0 22.1 57.1 20413 10813 10701 1 2 2 3 2.0 28.8 62.5 20501 11012 10902 1 1 7 4 7.0 18.9 59.3 20503 11012 10902 1 1 7 4 7.0 23.2 69.0 . . . . . . . . . . . . . . 41604 31514 31602 3 1 3 42 3.0 25.1 62.8 41612 31514 31602 3 2 3 42 3.0 23.3 66.6 41613 31514 31602 3 2 3 42 3.0 26.0 60.6 41614 31514 31602 3 2 3 42 3.0 27.0 65.2 41615 31514 31602 3 2 3 42 3.0 26.3 60.7

22. mouse.dat name of data fileMouse data from Karin Meyer [first of comment lines][This is mouseprp.in]mouse.dat [will now show file name in mtdf66] * end of comments7 number 'integer' fields3 number 'real' fields1 number of traits to analyzebody weight name of first trait2 position of trait in vector of 'reals'0. missing value indicator, cannot be '.'0 number of covariates 3 number of fixed factorsgeneration name of first fixed factor4 position in vector of 'integers'1 *** write summary of levels of factor, O would not writesex name of second fixed factor5 position in vector of 'integers'1 write summary

23. litter size name of third fixed factor6 position in vector of 'integers'1 write summary1 position of animal ID in vector of 'integers'329 *** number of animals (in pedigree from mtdfnrm, mtdf11, mtdf561 second animal effect (usually maternal genetic of dam)maternal genetic effect name for second animal effect 3 position of second animal effect in vector of 'integers'1 number of other random effectsmaternal permanent environment name of random effect3 position of random effect in vector of 'integers'0 *** don't write out summary, usually a big file1 *** write out file to match fixed factor levels with solutions 1 *** write out file to match other random levels with solutions

24. MTDF66 from MTDFPREP Started 07:05:16.51 on 08/10/1998 PROGRAM "MTDFPREP" - Setup W=X:Z matrix for MT-IAM Data set description : Mouse data from Karin Meyer single trait analysis mouse.ped No. of data lines in Unit 33 = 284 No. of integer variables per record = 7 No. of real variables per record = 3 No. of traits = 1 No. of valid records = 284 No. of animals in A-1 = 329 Order of MME (before constraints) = 712-------------------------------------------------------------------------------- Continued

25. Results for trait 1 - body weight (position 2 ) No. of records = 284 (missing value: .0000 No. missing = 0 )Trait Mean SD CV Min Max Std Min Std Max 1 24.0687 3.30236 13.72 14.600 34.500 -2.87 3.16 No. of covariates = 0 No. of fixed effects = 3 1: 3 levels for generation (MME rows: 1 - 3) Level Value No. % Mean 1 1 93 32.75 23.724 2 2 84 29.58 23.063 3 3 107 37.68 25.158 2: 2 levels for sex (MME rows: 4 - 5) Level Value No. % Mean 1 1 150 52.82 22.656 2 2 134 47.18 25.650 3: 7 levels for litter size (MME rows: 6 - 12) Level Value No. % Mean 1 1 11 3.87 26.609 2 2 41 14.44 23.722 7 7 30 10.56 21.973 No. of animals in A-1 = 329 (MME rows: 13 - 341) No. of 2nd animal effects = 1 1: 329 levels for maternal genetic eff (MME rows: 342 - 670) No. of uncorrelated random effects = 1 1: 42 levels for maternal perm env (MME rows: 671 - 712)-------------------------------------------------------------------------------- Continued

26. MTDF66 from MTDFPREP ContinuedOrder of MME (before constraints) = 712 Fixed effects = 3 Trait No. Name Position Levels Rows 1 1 generation 4 3 1 - 3 1 2 sex 5 2 4 - 5 1 3 litter size 6 7 6 - 12 Animal effects = 1 Trait No. Name Position Levels Rows 1 1 Animal w/ full A-1 1 329 13 - 341 2nd Animal effects = 1 Trait No. Name Position Levels Rows 1 1 maternal genetic eff 3 329 342 - 670 Uncorrelated random effects = 1Trait No. Name Position Levels Rows 1 1 maternal perm env 3 42 671 - 712-------------------------------------------------------------------------------- Files written: ∙ ∙ ∙

27. Mouse data from Karin Meyer {This file is 'mouserun.in']single trait analysis * end of comments0 *** for start or restart, 1 for continue or solns from previous run1 option to estimate variance components, for solns to MME use 40 number of constraints to be imposed by you0 *** not reordered previously, if reordered can use 1 if mtdf58 saved1 3.65354 starting value for a1,a12 .5359 starting value for m1,a1***3 1.37502 starting value for m1,m10 0 end input for G 1 yes, values on screen are correct0 number to be held constant1 .426651 starting value for c1,c1 other random factor0 0 end input for C1 yes, values on screen are correct 0 number to be held constant1 2.28714 starting value for e1,e1 residual variance0 0 end input for R1 yes, values on screen are correct 0 number to be held constant1 *** write solutions for covariates and fixed factors1 *** match original codes of fixed factor levels with solns1 *** write solutions for animal effects1 *** write solutions for other random factors1 *** match original codes of random factor levels1.d-6 convergence for variance of simplex -2logL300 number of simplex rounds

28. MTDF76 from MTDFRUN Started 07:05:23.98 on 08/10/1998++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ PROGRAM "MTDFRUN" - Estimate Covariance Components for MT-IAM++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Mouse data from Karin Meyer single trait analysis Cold start, i.e., not a continuation of previous run Run option 1: iterate for variance components 0 constraints imposed by user ****** (Co)variances in model: No. in likelihood calculation = 5 No. to be held constant = 0 No. to be maximized = 5 Starting values for this run: G matrix: 4.0 .5 .5 1.5 C matrix: .5 R matrix: 2.0 ** reordering called ** ** reordering completed ** The elapsed time was: 00:00:00.39 ** solve5 called ** ** solve5 completed ** The elapsed time was: 00:00:00.99Continue............

29. SPARSPAK-A statistics.. Time: Ordering = .270 secs. ( .004 mins.) Total/Solution = .170 secs. ( .003 mins.) Allocation = .000 secs. ( .000 mins.) Factorization = .109 secs. ( .002 mins.) Solve = .061 secs. ( .001 mins.) Storage: No. equations = 712 Non-zero hs elements in MME = 9742 Maximum storage required = 12304. ( .094 MB) Size of storage array (MAXSA) = 800000 ( 1.538% used) ******* RESULTS FROM SIMPLEX ******* OPTIONS SET FOR THIS RUN: MAXIMUM NO. OF SIMPLEX ITERATES ALLOWED = 300 MINIMUM VARIANCE OF FUNCTION VALUES IN SIMPLEX = .1000000000E-05 NO. OF SIMPLEX ITERATIONS CARRIED OUT = 44 NO. OF LIKELIHOODS EVALUATED = 79 NO. OF NON-PERMISSABLE PARAMETER VECTORS = 0 No. of failed contractions = 0 Variance of simplex function values = .2190791759E-06 Convergence criterion attained Continue............

30. MTDF76 continued Final Simplex: (++ best L; *** parameter held constant) 1 753.7488165 3.6828 .5114 1.3544 .4240 2.2586 2 753.7498457 3.6449 .4895 1.4246 .3877 2.2988++ 5 753.7486078 3.6407 .4745 1.3861 .4140 2.2980 6 753.7494189 3.7354 .5228 1.3221 .4393 2.2565 -2 log L = 753.7486077658 ( 5) Var = .0000002191 Half-store elements of AI inverse [Sample variances of estimates] 1 1 7.953967229489519 1 2 -3.619170958644206 1 5 -3.902998228002057 2 2 3.884627165149515 2 3 -3.275798374041433 2 5 1.821341783889839 3 3 6.128186867862764 3 4 -3.214987941640724 3 5 -1.224458463913176 4 4 2.092185282214288 4 5 5.792220185962393E-001 5 5 2.054178793146739

31. 0 Constraints: Estimates: GENETIC VARIANCES AND COVARIANCES : a1 m1 a1 : 3.64066 .47452 m1 : .47452 1.38608 UNCORRELATED RANDOM VARIANCES AND COVARIANCES: T1 C3 T1 C3 : .414033 ENVIRONMENTAL VARIANCES AND COVARIANCES : e1 e1 : 2.29804 PHENOTYPIC VARIANCES AND COVARIANCES : p1 p1 : 8.21333 HERITABILITIES AND GENETIC CORRELATIONS a1 m1 a1 : .44 ( .320) m1 : .21 .17 ( .712) ( .296) UNCORRELATED RANDOM EFFECTS T1 C3 T1 C3 : .50E-01 ( .177)

32. MTDF76 continuedENVIRONMENTAL PROPORTION OF TOTAL VARIANCE AND CORRELATIONS e1 e1 : .28 ( .196) Files written: MTDF4 (ascii): Parameter file (IUN5) for "cold" restart MTDF54 (ascii): Last simplex MTDF58 (binary): SPARSPAK reordering MTDF59 (ascii): Constraints imposed MTDF68 (ascii): Likelihoods by rounds MTDF67 (ascii): Sampling variances if requested MTDF72 (ascii): Predicted BVs and PEVs if requested MTDF76 (ascii): Program log file MTDF77 (ascii): Solutions for covariates and fixed effects if requested MTDF78 (ascii): Solutions for trait within animal if requested MTDF79 (ascii): Solutions for independent random effects if requested The current time is: 07:06:25.61 The elapsed time was: 00:01:00.64

33. Mouse data from K Meyer (This was mtdf4 copied to mtdf4.1 for restart) single trait analysis * end of comments 0 continuation: 0-no; 1-yes 1 run option 0 # constraints 0 reordered: 0-no; 1-yes <<< could change to 1 1 3.637814264727246 animal effect prior 2 5.289616129513786E-001 animal effect prior 3 1.300806074847860 animal effect prior 0 0.D0 1 values are correct 0 # parameters to hold constant 1 4.647326260772600E-001 uncorrelated effect prior 0 0.D0 1 values are correct 0 # parameters to hold constant 1 2.290805075973958 residual effect prior 0 0.D0 1 values are correct 0 # parameters to hold constant 1 write fixed effect solutions 1 merge PREP info with solutions 1 write animal solutions?? 1 write independent random effect solutions 1 merge PREP info with solutions 1.000000000000000E-006 convergence criterion <<< could change 300 # Simplex rounds <<< could change

34. MTDF76.2 after restart to check change in -2logL Started 14:10:31.49 on 08/26/2000++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ PROGRAM "MTDFRUN" - Estimate Covariance Components for MT-IAM Mouse data from Karin Meyer single trait analysis Cold start, i.e., not a continuation of previous run Run option 1: iterate for variance components Starting values for this run: G matrix: 3.6 .5 .5 1.3 C matrix: .5 R matrix: 2.3 ** reordering called and completed** MAXIMUM NO. OF SIMPLEX ITERATES ALLOWED = 300 MINIMUM VARIANCE OF FUNCTION VALUES IN SIMPLEX = .1000000000E-05 NO. OF SIMPLEX ITERATIONS CARRIED OUT = 15 NO. OF LIKELIHOODS EVALUATED = 34 Variance of simplex function values = .7619503856E-06 Convergence criterion attained Final Simplex: (++ best L; *** parameter held constant)++ 1 753.7462824 3.6378 .5290 1.3008 .4647 2.2908 6 753.7472056 3.6534 .5759 1.2561 .4875 2.2994 -2 log L = >>> 753.7462824159 <<< (1) Var = .0000007620 Large Sample Variances of Variance EstimatesContinued

35. MTDF76.2 continuedEstimates: GENETIC VARIANCES AND COVARIANCES : a1 m1 a1 : 3.63781 .52896 m1 : .52896 1.30081 UNCORRELATED RANDOM VARIANCES AND COVARIANCES: T1 C3 T1 C3 : .464733 ENVIRONMENTAL VARIANCES AND COVARIANCES : e1 e1 : 2.29081 PHENOTYPIC VARIANCES AND COVARIANCES : p1 p1 : 8.22312 HERITABILITIES AND GENETIC CORRELATIONS a1 m1 a1 : .44 ( .320) m1 : .24 .16 (1.137) ( .296) UNCORRELATED RANDOM EFFECTS T1 C3 T1 C3 : .57E-01 ( .178) ENVIRONMENTAL PROPORTION OF TOTAL VARIANCE AND CORRELATIONS e1 e1 : .28 ( .196) Files written: The elapsed time was: 00:00:01.54

36. MTDF77 [Fixed effect solutions] solution level * name trait ______________________________________________________ 1 -.16211010 1 * generation * body weight 2 -1.64614075 2 * generation * body weight 3 .01080553 3 * generation * body weight 4 -3.05158974 1 * sex * body weight 5 .00000000 <<<< 2 * sex * body weight 6 5.73575613 1 * litter size * body weight 7 2.22139506 2 * litter size * body weight 8 3.28421069 3 * litter size * body weight 9 1.34509621 4 * litter size * body weight 10 2.27238624 5 * litter size * body weight 11 1.53586287 6 * litter size * body weight 12 .00000000 <<<< 7 * litter size * body weight

37. MTDF79 [Uncorrelated random effect solutions] equn solution level * name trait_________________________________________________________ 671 -.04583612 10101 * maternal permanent e * body weight 672 .07454918 10404 * maternal permanent e * body weight 673 -.03016894 10506 * maternal permanent e * body weight 674 .03943667 10602 * maternal permanent e * body weight 675 -.08974688 10701 * maternal permanent e * body weight 676 -.09208224 10801 * maternal permanent e * body weight ∙ ∙ ∙ 704 -.09045377 30902 * maternal permanent e * body weight 705 -.18210848 30904 * maternal permanent e * body weight 706 -.00342698 31105 * maternal permanent e * body weight 707 .02015426 31203 * maternal permanent e * body weight 708 -.24623957 31302 * maternal permanent e * body weight 709 -.19488994 31304 * maternal permanent e * body weight 710 .13085314 31401 * maternal permanent e * body weight 711 .26078214 31501 * maternal permanent e * body weight 712 -.12766108 31602 * maternal permanent e * body weight

38. MTDF78 [Genetic solutions] Coded ID ID a1 m1___________________________________________________ 1 215 .77065064 .29387514 2 403 .77065064 .29387514 3 615 1.03862464 .49508782 4 701 1.03862464 .49508782 5 814 -.42773805 -.23195017 6 904 -.42773805 -.23195017 7 1314 -.39957873 -.16270164 8 1602 -.39957873 -.16270164 9 1701 .46445741 .46909996 10 1813 .46445741 .46909996 11 2201 .50407019 .56651447 12 2315 .50407019 .56651447 13 2401 -.12542643 -.11625806 14 2416 -1.52961347 -1.11182385 . . . . . . . . Continued

39. MTDF78 [Genetic solutions] continued ∙ ∙ ∙ 318 41512 .98956176 .40385134 319 41513 .28142180 .30088323 320 41516 2.22880667 .58404553 321 41517 1.96325419 .54543249 322 41601 .20518050 -.31031653 323 41602 1.04609669 -.18804190 324 41603 .24943925 -.30388102 325 41604 .86906171 -.21378393 326 41612 -1.27819113 -.52600831 327 41613 -.08320496 -.35224963 328 41614 .35938251 -.28789456 329 41615 .04957128 -.33294311 [For all animals in pedigree file ]

40. This answer file is 'mouserun.se' to do option 4 Note uses continuation option to read converged variance components Calculate F-stats, 'naive' t-stats for contrasts, and expectations * 1 continuation: 0-no; 1-yes [Files must be available]4 run option1 write fixed solutions1 match with prep ID0 do not write animal solutions0 do not write other random factor solutions1 calculate block of inverse and F-statistic4 first equation of block for sex5 last equation of block for sex1 calculate block of inverse for litter size6 first equation for block for litter size12 last equation for block for litter size 0 do not calculate another block of inverse of coef matrix

41. 1 calculate a contrast for sex 1 equn 4 vs sex 2 equn 52 number of elements in contrast4 1. equn and coef for sex 1 t stat squared = F 5 -1. equn and coef for sex 21 calculate contrast for generation 1 vs 22 number of elements in contrast1 1. equn and coef for generation 12 -1. equn and coef for generation 21 calculate contrast for generations 1 and 2 vs 33 number of elements in contrast1 1. equn and coef for generation 12 1. equn and coef for generation 23 -2. equn and coef for generation 30 no more contrasts

42. 0 no PEV and rTI1 calculate expectations of solutions1 equation corresponding to first parameter in expectation12 " " last " "1 equation corresponding to first solution in block for gen and sex5 " " last " "1 calculate more expectations1 first parameter in expectation12 last parameter in expectation6 first solution in block for litter size12 last solution in block0 no more expectations

43. MTDF67 from option 4 F-tests, t-tests, expectations Block of inverse for equations 4 to 5 4 .0608 .0000 5 .0000 .0000 inverse of inverse of order 2 with rank 1 F=statistic = 153.186 for eqn 4 to 5 with n = 2 but DF = 1 Block of inverse for equations 6 to 12 6 1.8511 .6086 .6306 .4512 .6347 .7711 .0000 7 .6086 1.2548 .6558 .8239 .8051 .7763 .0000 8 .6306 .6558 1.4760 .4720 .6430 .7432 .0000 9 .4512 .8239 .4720 1.4383 .7375 .6339 .0000 10 .6347 .8051 .6430 .7375 1.0013 .7082 .0000 11 .7711 .7763 .7432 .6339 .7082 1.3437 .0000 12 .0000 .0000 .0000 .0000 .0000 .0000 .0000 inverse of inverse of order 7 with rank 6 F=statistic = 3.617 for eqn 6 to 12 with n = 7 but DF = 6 Continued

44. Contrast 1 with no. elements 2 Row 4 coefficient 1.000000000000000 Row 5 coefficient -1.000000000000000 Contrast 1 =-.30515897E+01 SE = .24655668E+00 and t = -12.38 Contrast 2 with no. elements 2 Row 1 coefficient 1.000000000000000 Row 2 coefficient -1.000000000000000 Contrast 2 = .14840307E+01 SE = .64213778E+00 and t = 2.31 Contrast 3 with no. elements 3 Row 1 coefficient 1.000000000000000 Row 2 coefficient 1.000000000000000 Row 3 coefficient -2.000000000000000 Contrast 3 =-.18298619E+01 SE = .11472325E+01 and t = -1.60Continued

45. MTDF67 from option 4 F-tests, t-tests, expectations continuedContributions of parameters 1 to 12 To expectations for solutions 1 to 5 Parameter: Coefficients for solutions 1 to 5 1 1.000 .000 .000 .000 .000 2 .000 1.000 .000 .000 .000 3 .000 .000 1.000 .000 .000 4 .000 .000 .000 1.000 .000 5 1.000 1.000 1.000 -1.000 .000 6 .000 .000 .000 .000 .000 10 .000 .000 .000 .000 .000 11 .000 .000 .000 .000 .000 12 1.000 1.000 1.000 .000 .000

46. Contributions of parameters 1 to 12 To expectations for solutions 6 to 12 Parameter: Coefficients for solutions 6 to 12 1 .000 .000 .000 .000 .000 .000 .000 5 .000 .000 .000 .000 .000 .000 .000 6 1.000 .000 .000 .000 .000 .000 .000 7 .000 1.000 .000 .000 .000 .000 .000 8 .000 .000 1.000 .000 .000 .000 .000 9 .000 .000 .000 1.000 .000 .000 .000 10 .000 .000 .000 .000 1.000 .000 .000 11 .000 .000 .000 .000 .000 1.000 .000 12 -1.000 -1.000 -1.000 -1.000 -1.000 -1.000 .000

47. mouse.dat Mouse data from Karin Meyer (This file is 'mouseprp.in2'two trait analysis* end of comments7 number 'integer' fields3 number 'real' fields2 number of traits to analyzebody weight name of first trait2 position of trait in vector of 'reals'0. missing value indicator, cannot be '.'0 number of covariates 3 number of fixed factorsgeneration name of first fixed factor4 position in vector of 'integers'1 write summary of levels of factor, O would not writesex name of second fixed factor5 position in vector of 'integers'1 write summarylitter size name of third fixed factor6 position in vector of 'integers'Continued

48. litter size name of third fixed factor6 position in vector of 'integers'1 write summary1 postion of animal ID in vector of 'integers'329 number of animals (in pedigree from mtdfnrm, mtdf11, mtdf561 second animal effect (usually maternal genetic of dam)maternal genetic effect name for second animal effect 3 position of second animal effect in vector of 'integers'1 number of other random effectsmaternal permanent environment name of random effect3 position of random effect in vector of 'integers'0 don't write out summary, usually a big fileContinued

49. mouse.dat Mouse data from Karin Meyer (This file is 'mouseprp.in2'two trait analysis Continued)other trait3 position0. missing value indicator0 covariates3 fixed factorsgeneration 4 position1 write summarysex 5 position1 write summarylit size6 position1 write summary1 second animal effectmat gen 2 (name)3 position1 other random factor mat pe (name)3 position0 do not write summary1 write file to match fixed factor levels with solutions 1 write file to match other random levels with solutions

50. Mouse data from Karin Meyer (This file is 'mouserun.in2')two trait analysis* end of comments0 for start or restart, 1 for continue or solns from previous run1 option to estimate variance components, for solns to MME use 40 number of constraints to be imposed by you0 not reordered previously,if reordered can use 1 if mtdf58 saved1 4. starting value for a1,a12 3.5 a1,a23 .5 starting value for m1,a14 .4 a1,m25 20. a2,a26 .1 a2,m17 2. a2,m28 1.5 starting value for m1,m19 .1 a1,m210 7. m2,m20 0 end input for G 1 yes, values on screen are correct0 number to be held constant

51. 1 .5 starting value for c1,c1 other random factor2 .5 c1,c23 2. c2,c20 0 end input for C1 yes, values on screen are correct 0 number to be held constant1 2. starting value for e1,e1 residual variance2 .2 e1,e23 10. e2,e20 0 end input for R1 yes, values on screen are correct 0 number to be held constant1 write solutions for covariates and fixed factors1 match original codes of fixed factor levels with solutions1 write slutins for animal effects1 write solutions for other random factors1 match original codes of random factor levels1.d-6 convergence for variance of simplex -2logL100 number of simplex rounds

52. This is answer file 'mouserun.se2' for 2-traits with option 4 Continuation option uses file in memory of converged VC Calculates 'naive' SE of contrasts, specified SEP and rTI, and expectns * end comments 1 continuation: 0-no; 1-yes4 run option1 write fixed solutions1 match with prep ID1 do write animal slutions0 do not write other random factor solutions0 do not calculate block of inverse of coef matrix1 calculate a contrast for sex 1 equn 4 vs sex 2 equn 52 number of elements in contrast4 1. equn and coef for sex 15 -1. equn and coef for sex 21 calculate contrast for generation 1 vs 22 number of elements in contrast1 1. equn and coef for generation 12 -1. equn and coef for generation 21 calculate contrast for generations 1 and 2 vs 33 number of elements in contrast1 1. equn and coef for generation 12 1. equn and coef for generation 23 -2. equn and coef for generation 30 no more contrasts

53. 1 PEV and rTI25 first animal equn30 last animal equn 1 first genetic trait 4 last genetic trait (a1,a2,m1,m2) Does not ask for other PEV?? 1 calculate expectations of solutions 1 equation corresponding to first parameter in expectation12 " " last " " 1 equation correspondingto first solution in block 5 " " last " " 1 calculate more expectations 1 first parameter in expectation12 last parameter in expectation 6 first solution in block12 last solution in block 0 no more expectations

54. MTDFNRM MTDFPREPpedigree file data filefree format free format MTDF11MTDF44MTDF13MTDF56MTDFRUNMTDF21MTDF22MTDF50MTDF51MTDF52MTDF66MTDF4MTDF54MTDF58MTDF59MTDF68MTDF72MTDF76MTDF77MTDF78MTDF79MTDF67CONTINUERESTART••••••