Hi,

I would say that you can attempt to reconstruct phased haplotypes from unphased genotypes and then estimate LD, using software packages such as fastPHASE (Scheet and Stephens 2006), BEAGLE (Browning and Browning 2007), and MaCH (Li et al. 2010). For putative double heterozygotes, these packages infer haplotypes from relative frequencies of putatively unambiguous genotypes, which are themselves called using another software package. Therefore, these software packages do not maximally use the information for LD estimation, in particular on haplotype phase, contained in the sequence-read data (Bansal et al. 2008; Long et al. 2009)

You can also use TASSEL software.

if you have genotype data, you can try PLINK. If you only have a list of SNPs, you can try SNAP.

https://www.broadinstitute.org/mpg/snap/ldsearchpw.php

http://pngu.mgh.harvard.edu/~purcell/plink/ld.shtml