You could use snp_sites to find polymorphic sites (SNPs) in your whole genome alignments https://github.com/sanger-pathogens/snp_sites.

I have created a basic analysis pipeline (http://mcgp.sourceforge.net/) that create alignments based on the whole genome content that's conserved between all strains and identifies all variation (SNPs) using the inferred 'core genome'. The could later be used to create phylogenetic trees at species level for 'bacterial' species.

If you have assemblies already, all you would need to do is order them using abacas to create one pseudo genome for each isolates. These ordered sequences can be provided as inputs for the script.