c c Robust Version of the WPC method: 6 december 1995, revise 18 decembre 1996. c c c Genetic linkage test by the WPC method. c This program implements the WPC method: the variance of the score statistic c is computed by exact formulae from the permutation distribtution. c This present method takes into account that there is an c intrasibship correlation. This correlation can be estimated, then, c a transformation is done to ensure exchangeability c of the transformed residuals c which is a necessary assumption for using permutations. c c This method is presented in : c Commenges, D and Abel L. (1996) Improving the robustness of the WPC test c for linkage analysis. Genet Epidemiol, 13. c Previous publications: c Commenges, D. (1994), Genet Epidemiol 11: 189-200; c Commenges, D., Olson, J. and Wijsman, E. (1994) c Genet Epidemiol 11: 201-212; c Commenges, IGES, Paris, 1-2 June, 1994 (see Genet Epid 11, 290). c Commenges, D. (1995), Genet Epidemiol 12: 853-57. c c------------------------------------------------------------ c USER'S MANUAL c------------------------------------------------------------ c This program allows to compute three variants of WPC: c WPC-MP and WPC-RP (previously called WRPC) for survival data c and WPC-OP for quantitative traits. c The main program reads parameters in the file 'parwpc' c the detailed results will be written in the file 'detoutwpc' c and summary results will be written in the file 'sumoutwpc' c In 'sumoutwpc', the score, its variance and the test statistic WPC c are given for the whole set of families and for each marker. c In 'detoutwpc', the results are given for each family. c There is a first loop on the families for computing c the number of subjects in each family. c Then, there is a second loop to estimate a global intrasibship correlation. c (this loop is skipped if the initial value of ro is greater or equal to 0. c The third loop on the families is for computing the WPC statistics c for up to 15 markers. c For each family the main program selects n subjects c after eliminating some subjects; c Then it calls the subroutine mulwpc which computes the values c of the chosen wpc statistic. c At the end of the loop on the families, the program computes c the global statistics for the set of families: WPCt. c c USAGE: c c wpc # without arg's wpc use 'parwpc', 'detoutwpc', 'sumoutwpc' files c c wpc parwpc-file detoutwpc-file sumoutwpc-file c c Mailto: Daniel.Commenges@bordeaux.inserm.fr c c Detailed description of the input files c---------------------------------------------- c The parameter file 'parwpc': c---------------------------------------------- c first line: name of the file containing the phenotypes (maximum 128 letters) c c second line: reading format for this file; if you put free, the reading c format will be free, that is data are separated by a blank or a coma. c c third line: name of the file containing the genotypes (maximum 128 letters) c c fourth line: reading format for this file; if you put free, the reading c format will be free, that is data are separated by a blank or a coma. c c fifth line: maximum number of markers treated; this number must c be lower or equal to the number of markers in the genotype file c c sixth line: method, missing value, ro , seuil c method=1: quantitative trait (this can also treat a binary trait): c WPC-OP c method=2: survival data (age, status): WPC-MP c method=3: survival data, rank residuals: WPC-RP (WRPC) c missing value bound: all values equal to that value for integers c (status, age) or lower or equal for real numbers(quantitative trait) c will be considered as missing. c In addition, age and alleles codes will be considered as missing c if their value is lower or equal to zero. c If age or status are missing for WPC-MP c or WPC-RP subjects will be eliminated; if the quantitative trait is missing c and the method is WPC-OP, the subject will be eliminated.; c ro (intrasibship correlation): if ro is greater or equal to 0, c this value will used c if ro<0, ro will be estimated; c seuil: if ,for one family, WPC > seuil the value c of WPC will be written on 'detoutwpc'. c c seventh line: this is a grid containing numbers equal to 0 or 1 separated by c blanks. c the number of 0 or 1 is given in the fifth line. c those markers for which a 1 have been put will be treated. c c It is possible to make several requests in this file; c for instance you can repeat the same lines, just changing the name of the c genotype file for treating another chromosome. c c-------------------------------------------------- c The phenotype file: c-------------------------------------------------- c for each subject it must contain (on one line) in the given order: c family ID, subject ID, father ID, mother ID, sex, age, c disease status, quantitative trait c All these variables are integers except the quantitative trait c which is real. (if the format is free, only these variables, separated c by a blank, must be on the record; there may be other variables c after the quantitative trait which will not be read. c If you give a format, you can skip c some variables.) c Undefined values: see miising value in the parwpc file. c for method 2 and 3 subjects with undefined values c for age or status will be eliminated c for method 1 subjects with undefined values for quantitative trait c are eliminated c c---------------------------------------------------- c The genotype file: c---------------------------------------------------- c for each subject it must contain (on one line) in this order: c family ID, subject ID, allele1 (marqueur 1), allele2 (marqueur 1), c allele1 (marqueur 2), allele2 (marqueur 2), ..., c allele1 (marqueur mkm), allele2 (marqueur mkm) c subjects are eliminated if at least one allele code for one c of the loci specified by the grid is lower or equal to zero. c c---------------------------------------------------- c Present limits: c---------------------------------------------------- c number of families: nbfam=300; c number of subjects by family: num=200; c total number of subjects in the set of families: 10000; c total number of sibships: 1000; c number of alleles: 20. c number of markers treated in one pass: 15 c c (end-of-manual)