Item

Abstract

Background: Segmental duplications (SDs) are long DNA sequences that are repeated in a genome and have high
sequence identity. In contrast to repetitive elements they are often unique and only sometimes have multiple copies in
a genome. There are several well-studied mechanisms responsible for segmental duplications: non-allelic homologous
recombination, non-homologous end joining and replication slippage. Such duplications play an important role in
evolution, however, we do not have a full understanding of the dynamic properties of the duplication process.
Results: We study segmental duplications through a graph representation where nodes represent genomic regions
and edges represent duplications between them. The resulting network (the SD network) is quite complex and has
distinct features which allow us to make inference on the evolution of segmantal duplications. We come up with the
network growth model that explains features of the SD network thus giving us insights on dynamics of segmental
duplications in the human genome. Based on our analysis of genomes of other species the network growth model
seems to be applicable for multiple mammalian genomes.
Conclusions: Our analysis suggests that duplication rates of genomic loci grow linearly with the number of copies of
a duplicated region. Several scenarios explaining such a preferential duplication rates were suggested.