[ < ] [ > ]   [ << ] [ Up ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4. Computing Kinship and One- or Two-Locus Inbreeding Coefficients

4.1 Introduction to kin  
4.2 Sample kin parameter file  
4.3 Running kin example and sample output  
4.4 kin statements  


[ < ] [ > ]   [ << ] [ Up ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4.1 Introduction to kin

kin computes kinship coefficients for pairs of pedigree members. It also computes single-locus and two-locus inbreeding coefficients for members of the pedigree. Briefly, the kinship coefficient between individuals i and j is the probability that a randomly-drawn allele from i is identical by descent (ibd) to randomly-drawn allele from individual j at the same locus. A single-locus inbreeding coefficient is the probability that an individual carries two copies of a gene that are ibd, at a given autosomal locus. In other words, an individual's single-locus inbreeding coefficient is equal to the kinship coefficient of his parents, as an individual's gametes can be thought of as random draws from his parents' chromosomes. A two-locus inbreeding coefficient is the probability that an individual carries two ibd copies of a gene at each of two linked loci. kin presents two-locus inbreeding coefficients as a function of the recombination fraction between the two loci.

Note: The kin program does check for duplicate requests within a pedigree component of any inbreeding or kinship coefficients; each will be computed once only, even if the individuals are specified in reverse order. However, it does check (and quits with an error) if a request is made for kinship of an individual with him/her self. This bug will be fixed in a future MORGAN release.

See Concept Index for: kin introduction, kinship coefficient, inbreeding coefficient, ibd,


[ < ] [ > ]   [ << ] [ Up ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4.2 Sample kin parameter file

Files for kin may be found in the `IBD' subdirectory of `MORGAN_Examples'. Below is a sample kin parameter file, `jv_rep_kin.par'.

 
set printlevel 5
input pedigree file 'jv_rep.ped'
compute component 1 kinship coeff            531 431   431 432
compute component 1 inbreeding coeff         332   531
compute component 2 kinship coeff            341  442
compute component 2 inbreeding coeff         441   541
compute component 1 two-locus inbreed coeff  531
compute component 2 two-locus inbreed coeff  441
set recomb freqs .01 .05 .04 .10 .18 .30 .50 .0

The statements on lines 3 -- 8 request computation of kinship coefficients for the pairs `531 431' and `431 432', and then inbreeding coefficients for individuals `332' and `531', from component 1. It then requests kinship coefficients for the pair `341 442' and inbreeding coefficients for individuals `441' and `541' from component 2. Finally, it requests the two-locus inbreeding coefficient for `531' from component 1 and `441' from component 2. The two-locus inbreeding coefficient will be computed for two loci at distances specified in the `set recomb freqs' statement. (Note these need not be ordered, but the program will order them in the output.) If there is more than one component (connected pedigree) in the file, the component number must be specified. MORGAN assigns component numbers to the connected pedigrees within the pedigree file. If your data set contains more than one component, you may first run pedcheck to determine which individuals are assigned which component numbers.

See Concept Index for: kin sample parameter file.


[ < ] [ > ]   [ << ] [ Up ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4.3 Running kin example and sample output

Under the subdirectory `IBD/', run the example above using the command below. To send the output to a file instead of the screen, include `> filename' (without quotes) after the parameter file name on the command line:

 
./kin jv_rep_kin.par    or
./kin jv_rep_kin.par > out-file-name   

Below is the relevant part of the kin output.

 
 Component 1:

    Kinship coefficients:

    531  431   .32031
    431  432   .10938


    Inbreeding coefficients:

    332   .00000
    531   .10938


    2-locus inbreeding coefficients:
    (g4link is probability of IBD at both of 2 linked loci)

    proband  recomb   g4link
               freq     prob

        531    .000   .10938
               .010   .10234
               .040   .08386
               .050   .07849
               .100   .05660
               .180   .03455
               .300   .01910
               .500   .01196


 Component 2:

    Kinship coefficients:

    341  442   .15625


    Inbreeding coefficients:

    441   .06250
    541   .10938


    2-locus inbreeding coefficients:
    (g4link is probability of IBD at both of 2 linked loci)

    proband  recomb   g4link
               freq     prob

        441    .000   .06250
               .010   .05885
               .040   .04905
               .050   .04614
               .100   .03388
               .180   .02060
               .300   .01008
               .500   .00391

Note that when the recombination frequency is 0.0, the two-locus inbreeding coefficient is the same as the one-locus inbreeding coefficient, as there is no recombination between the loci, thus they act as a single locus. When the recombination frequency is 0.5, the two loci are independent and the two-locus inbreeding coefficient is the square of the one-locus inbreeding coefficient.

See Concept Index for: running kin example, kin sample output.


[ < ] [ > ]   [ << ] [ Up ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4.4 kin statements

At least one of the following `compute ...' statements are required to run program kin. If there is more than one component (connected pedigree) in the file, the component number must be specified. MORGAN assigns component numbers to the connected pedigrees within the pedigree file. If your data set contains more than one component, you may first run pedcheck to determine which individuals are assigned which component numbers. pedcheck will sort by component number in the output pedigree file, although it will not list component numbers in the file. The screen output generated when running pedcheck will give component numbers and the number of individuals in each component. Check the error and warning messages when running kin to verify that component numbers were correctly specified. The program will quit if an individual's component number is incorrectly specified in the parameter file or if there is more than one component in the data set and no component is specified.

compute [component M] kinship coefficient N1 N2...

This statement names one or more pairs of pedigree members for which the kinship coefficient is to be computed.

compute [component M] inbreeding coefficient N1...

This statement names one or more pedigree members for whom the inbreeding coefficient is to be computed.

compute [component M] two-locus inbreeding coefficient N1...

This statement requests the computation of two-locus inbreeding coefficients, i.e. the probability of ibd at both loci, for the named individual. For the recombination frequencies at which the coefficients are computed, see the following statement.

set recombination frequencies X1 X2...

Two-locus inbreeding coefficients are computed for each of the list of recombination frequencies, in the range of 0.0 to 0.5. If frequencies are not given, the default values are: 0.0, 0.01, 0.04, 0.05, 0.10, 0.18, 0.30, and 0.50.

See Concept Index for: kin statements, component, pedigree component, kinship coefficient, inbreeding coefficient, recombination.


[ << ] [ >> ]           [Top] [Contents] [Index] [ ? ]

This document was generated by Elizabeth Thompson on July, 7 2013 using texi2html