ldgroup: 
linkage disequilibrium grouping of single nucleotide
polymorphisms (SNPs) reflecting haplotype phylogeny
for efficient selection of tag SNPs.

### URL, CONTACT
http://www.fumihiko.takeuchi.name/publications.html
Questions, comments or bugs?  Please write to the address mentioned in
the above web page, quoting 'ldgroup' in title.

### REFERENCE
The program ldgroup implements the algorithm explained in the
following paper.  Please cite it for reference.

Takeuchi, F. et al. (2005)
Linkage disequilibrium grouping of single nucleotide polymorphisms (SNPs)
reflecting haplotype phylogeny for efficient selection of tag SNPs,
Genetics 170, 291--304.

### PROGRAM FILE
README.txt   this file
ldgroup.R   the program written in R

### R
The ldgroup program is written in R language, which is available from
http://www.r-project.org/

The program also requires an R library 'igraph', which can be
downloaded using your R program.
Relevant information is available from the following URL.
http://www.stats.bris.ac.uk/R/src/contrib/Descriptions/igraph.html
http://cneurocvs.rmki.kfki.hu/igraph/

### INPUT FILE
Differently from the explanation in the paper, ldgroup uses 'phased'
genotype (ie haplotypes) as input.  Thus, you need to phase your
genotype data before running ldgroup.  Since, phasing programs usually
impute missing genotype, the input for ldgroup assumes to include no
missing data.  Output format of two phasing programs are readily supported.

One is the Beagle software by Browning et al:
http://www.stat.auckland.ac.nz/~browning/beagle/beagle.html
I also prepared a sample input file for ldgroup in this format:
test.phased

Another is the MACH software by Li et al:
http://www.sph.umich.edu/csg/abecasis/MACH/index.html
I also prepared sample input files for ldgroup in this format:
test.geno
test.dat

### OUTPUT FILE; TAG SNP SELECTION
Output file names are filenameradix.tagSNPs<s>.txt
where <s> is the r^2 threshold for defining LD groups

There is one row for each SNP.  The SNP ID is in the third column.
The SNPs are roughly classified by LD groups (first column), and then
subdivided into complete-LD subgroups (second column).  The SNPs
having the same number for the first and second columns belong to the
same complete-LD subgroup, and could be regarded as equivalent.
Selecting one each from the equivalent SNPs comprises a set of tag
SNPs.

### VERSION HISTORY
2007.11.16

### LICENSING
GNU General Public License
(If you are not happy with this, please contact me for other licencing.)