IBD documentation
/* this is the ibd program version 0.1 which can handle multiple markers. */
This program is written to partially replace the SIBPAL program.
Given data on a set of nuclear families, and appropriate gene frequencies, this
program will perform the following functions:
(1) estimate, for each sib pair and each locus, the proportion (pi) of
genes identical by descent that the sib pair shares at that locus;
(2) give the sib-pair trait value for each trait.
The source codes are
ibd.c, getlh.c
The include files are
ibd.h, ibdfunc.h, getlh.h
To make the executable file in Unix, at the command prompt type:
make
The executable file named:
ibd
To run the program, make sure both files
fort.1 the parameter file, prepared by user
fort.12 the nuclear family data file
are in the directory when the command 'ibd' is issued.
By default, the output goes to file
loci.out (f1,f2 and pi values for multiple marker loci)
trait.out (sib-pair trait values for multiple traits)
ibd.log (log file for missing value report and summary of 'fort.12')
(1). fort.1 --- parameter file:
"fort.1" is the input parameter file for the IBD program and a description
of the format follows: ( A: character, F: real number, I: integer)
Record 01: Problem Identification
Columns Format Information contained
------- ------ ---------------------
1-80 A80 Title of this run.
Record 02: Printing options
(still under revision; any input is o.k.)
Record 03: Type of Analysis
Columns Format Information contained
------- ------ ---------------------
1-5 I5 Number of traits.
6-10 I5 Number of marker loci.
11-15 I5 Number of covariates.
Record 04: Traits
(Repeat for each trait.)
Columns Format Information contained
------- ------ ---------------------
1-10 A10 Name of the trait.
11-15 F5.0 Missing value for the trait.
Record 05: Marker Loci
(Repeat for each marker locus.)
Columns Format Information contained
------- ------ ---------------------
1-10 A10 Name of the marker locus.
11-15 F5.0 Missing value for the marker locus.
Record 06: Covariates
(Repeat this record for each covariate.)
Columns Format Information contained
------- ------ ---------------------
1-10 A10 Name of the covariate.
11-12 A2 Missing value for the covariate.
Record 07: Format For Family Data File Records
(The format can take up to two 80-character lines. If only one line
is used, leave the second line blank.)
Columns Format Information contained
------- ------ ---------------------
1-80 A80 Format to read the family data file.
The format must start with a
left-parenthesis, and end with
right-parenthesis.
The information must follow the
sequence indicated below:
1. Traits
(F-format:real number)
2. Marker Loci
(A-format:character, where
a2 or a3 means genotype,
a4,a5 ... means phenotype)
3. Covariates (if any)
(F-format:real number)
Record 08-13:
(Save for other program.)
Record 14: Locus Description Records
(Same as 4.3 Locus Description Records (page 18)
in SIBPAL v2.5)
(2). fort.12 --- data file:
If you are analyzing your own family data, you would have to prepare this file into
the standard format described here. The data should be nuclear family data, with a
header part running at the beginning. For the sample provided, the header indicates
that the data has 40 families with the size 6 of each one. After header part, there
must be the individual data records for each family in order. Each record contains the
following information ( does not need to be in the certain order):
family_id,
individual_id,
father_id,
mother_id,
sex(1=male, 2=female),
affection_code (1=unaffected, 2=affected),
markers (variable number allowed, example 'fort.12' has 3 of them)
trait
undefined (to be implemented)
Within each family, parents will take up the first two lines. The parents
do not have their father id and mother id, so the space is either left blank
or 0.
=========== END OF README ==============
Written by L. Emily Yu; July 10, 1996.