The Illumina HiSeq2000 paired-end sequence data were analyzed with BWA alignment software to map each read pair onto a genome reference.

We aligned our sequence data to either Hg18 (NCBI build 36.1) or Hg19 (GRCh37) human reference genome. 
For Hg18 (NCBI build 36.1) alignments, the genomic sequence was restricted to chromosomes 1-22, M, X and Y. 
For Hg19 (GRCh37) alignments, this reference contained genomic sequences for all chromosomes in the GRCh37 reference genome used for the 1000 genome project.

BWA's aln and sampe tools were used to generate small-gapped global alignments on the paired-end reads [ref 1].
BWA with the default parameter settings were used for alignments.

In the final stage of the alignment, we set the record's flags. First, we took all reads that failed Illumina's Chastity filter and turned on bit 512 in that record's bitwise flag to indicate the read failed platform/vendor quality checks. Then we ran Picard's MarkDuplicates program to set bit 1024 in the bitwise flag.

[ref 1] Li H. and Durbin R. Fast and accurate short read alignment with Burrows-Wheeler Transform. Bioinformatics 2009; 25:1754-60.