Our final release data sets which can be found in the release directory tend to be fully phased and useable in imputation experiments. The phase status of all our snps in vcf files should be defined by the delimiter of the genotype, | means the genotype is phased / means the genotype is unphased. For more information about the vcf format please look at the vcf 4.1 specification.

Release files do contain R squared values for all the phased genotypes and you can generally see subsets of this data in the release supporting directory.

Some individual genotypes in our variant call sets are derived by imputation because there is no coverage in a particular location for that individual.

For many releases standard files were created for the imputation tools Mach, Beagle and Impute2. There is more information about our most recent release available in the README.20111111_phase1_integrated_call_set