Similarity of Integration Sites of Different Integration Complexes

This report studies data created by Mary Lewinski and others in and associated with the laboratory of Frederic Bushman at the University of Pennsylvania. The aim of this report is to assess the tendency of different integration complexes — in this case retroviral vectors composed of HIV, MLV, or elements of both — to select particular genomic loci as favored integration targets. Previously, it has been shown that HIV and MLV favor different sites for integration. It is of particular interest to characterize the degree to which different integration complexes favor the same or different sites. With a very large number of integration events (of the order of 10 per base by complex or 150,000,000,000 for this study), this could be done directly by counting the number of events at each genomic locus for each integration complex, then comparing the counts. Integration complexes that tend to share high counts at some loci and share low counts at other loci presumably share features that govern integration targetting. On the other hand, integration complexes whose counts do not correlate in this fashion presumably do not share features relevant to integration targetting.