There will be a few, but these might also be sequencing artefacts. You never know until you validate. If you want to play around with real twin data (generated by Complete Genomics), you can do that here: http://genomecomb.sourceforge.net/publications/filters/filterselection.php You can customize the filter properties.Depending on how stringent you are you'll get less false positive differences and more true positive real differences between the two genomes.
Disclaimer: data and tool from my department
Most research suggests none: