Program rev3_4 converts Revision 3 input files peel.dat, phen.dat, and marker.dat to their Revision 4 equivalents. Files trip.dat and order.dat replace peel.dat. See sections II.1 and II.2 for the contents and sections B.1 and B.2 for the formats of trip.dat and order.dat. Files header.dat, phen.dat and ascer.dat replace Revision 3 phen.dat. See section II.3 and II.4 for the contents and section B.3 and B.4 for the formats of header.dat, phen.dat and ascer.dat. File popln.dat replaces marker.dat. See section II.5 for the contents and section B.5 for the format of popln.dat.
File papin.dat replaces pelphn.dat; produce a new version by executing preped. See section III.1 for a description of preped. File model.dat differs completely; produce a new version by executing prepap. See section III.3 for a description prepap.
Output file pap.out replaces lods.dat and pap.log is an added output file used for the simulation options.
Program pedlis and phnlis have been eliminate; programs descstat, simul, and gpe have been added. See section III.2, III.4, and III.6, respectively, for descriptions of descstat, simul, and gpe.
Program papdr now has eight rather than five execution options which differ from Revision 3 as follow:
(1) Single likelihood. The output from papdr now includes the pedigree number. Section IV.1.1 describes execution option 1.
(2) Genotypic probabilities. The selection to compute probabilities on all pedigree members can include all or only measured individuals. Section IV.1.2 describes execution option 2.
(3) Simulation and write files. This is a new option. Section IV.1.3 describes execution option 3.
(4) Maximization (previously option 3). NPSOL (purchased from the Department of Operations Research, Stanford University, Stanford, CA 94306) is an alternative to GEMINI. Section IV.1.4 describes execution option 4.
(5) Standard errors (previously option 4). Section IV.1.5 describes execution option 5.
(6) Simulation and estimation. This is a new option. Section IV.1.6 describes execution option 6.
(7) Expected lod scores. This is a new option. Section IV.1.7 describes execution option 7.
(8) Grid (previously option 5). Section IV.1.8 describes execution option 8.
See section III.5 for more information on papdr.
Sections IV.2 through IV.9 describe in detail the models that are available in Revision 4. The major change is that rather than selecting a penetrance subroutine, you now select subroutines for the major locus (for discrete and quantitative traits separately if both are included in the model) and a subroutine for the within-genotype model. Any combination can be used which allows multivariate models comprising a mixture of discrete and quantitative traits.
The mixed model approximations available in Revision 3 [Hasstedt 1982, 1991] have been replaced by a new approximation [Hasstedt 1993]. The corresponding use of a new algorithm allows calculation of the likelihood and simulation of phenotypes for diverse types of multivariate models. See section VI.10 for more information.
The option in Revision 3 to consider specified marker alleles as indistinguishable as a way to speed computation has been eliminated in Revision 4. An algorithm has been added to Revision 4 which identifies possible genotypes in untyped pedigree members and equates genotypes for which the alleles do not occur in any descendants. See section VI.3 for more information on this algorithm.
In Revision 3 the user specified the computational order of the nuclear families in peel.dat. In Revision 4, an algorithm determine computational order of the nuclear families unless the user specifies an order in order.dat. See section VI.1 for more information.