In collaboration with the group of Yannis Almirantis at NCSR "Demokritos" we return to the concept of Chargaff's 2nd parity rule of nucleotide composition and deviations thereof.
In a novel and rather exhaustive analysis conducted by Konstantinos Apostolou-Karampelis as part of his PhD Thesis, we were able to show how a straight-forward analysis of nucleotide composition can reveal very significant aspects of genome evolution and the mutational mechanisms underlying it. Konstantinos started by analyzing more than 300 bacterial genomes, comparing the preference of specific dinucleotides occurring with different frequencies between the two DNA strands. Through a simple measure of this asymmetry, he showed that differences in these preferences can trace the phylogeny of bacteria in a way comparable to standard genome alignment methods. He then went on to show how these differences may be attributed to the different mutational biases towards different GC contents in bacteria. More importantly he showed that these bias are coupled to the structure of the genetic code and that the combination of a particular GC bias can lead to striking sequence composition through the constraints imposed by the genetic code.
Last but not least, this work is the first to provide a mechanistic link between the observed genome composition patterns and molecular processes. By linking the pattern of each species to its employment of different DNA polymerase II-a subunit, we were able to associate specific pattern groups with different enzyme isoforms existing in diverse phyla.
This is thus the first time the enzymatic activity of DNA polymerases is directly linked to the genome sequence in bacteria.
For those brave enough to go into the details of the model and the analysis you can find the full paper here: