Biomedical Engineering Reference
In-Depth Information
In general, the elimination of genomes from the population because of stop codon
appearance is more frequent than their elimination by the loss of the coding signal. The
elimination because of the coding signal is additionally delayed in the simulations, which
indicates that gene sequences need some time to accumulate sufficient number of sub-
stitutions to change their coding properties. It seems to be important to notice, that the
effect of selection pressure used in all simulations was exclusively negative. The higher
robustness of coding sequences with enhanced coding signal after substitutions could
be considered as some hidden no direct positive selection. Nevertheless, it would be in-
teresting to introduce the direct positive selection effect of the increased coding signal.
The obtained results indicate that it is not indifferent to genes from the differently
replicated DNA strands to which mutational pressure and selection they are subjected.
The leading strand mutational pressure is more destructive for coding signal than the
lagging strand pressure. Therefore it is more harmful for genes from the both DNA
strands. On the other hand, the pressure typical of the lagging strand eliminates more
genes because of stop codons occurrence than the leading strand pressure because the
lagging strand substitutions generate such codons with higher frequency. The results
are in agreement with analysis of nucleotide usage biases in four-fold degenerated sites
in codons from bacterial genes [36].
Interestingly, the reverse and changing pressures destroy the coding signal weaker
than the direct pressure. As a result of this, the number of eliminated genes subjected
to these pressures is smaller than for the direct pressure and the average number of
accumulated substitutions shows some excess for the simulations with the changing
pressure. These findings agree with other simulations of gene evolution using selection
constraints on amino acid composition [21-24]. They also showed that genes are less
frequently eliminated when they change their mutational pressure between the differ-
ently replicated DNA strands periodically. These results of computer simulations are in
agreement with the comparative genome analysis showing that genes very often change
their position between the different strands [26] and homologs located on these strands
evolve faster than homologs located in the same type of DNA strand [4, 18-20].
In contrast to that, generally more genes are eliminated under the reverse and chang-
ing pressures when the selection against stop codons is considered although the dif-
ference to the direct pressure is not very big. However, if the location of the genes on
the DNA strands is taking into account, the lagging strand genes under the changing
pressure are the most often eliminated whereas the leading strand genes are the least
frequently removed by this type of selection of all possible scenarios. The applied pe-
riod of changing the mutational matrix equal to
0
.
5
million MCS was chosen arbitrarily
and probably there exist more optimal conditions for different genes as they were found
in other type of simulations [21].
We expect that the obtained results of simulations should be very similar to those
using other selection algorithms predicting protein coding sequences based on other
coding measures, e.g. codon or dicodon usage, because the measures used by the al-
gorithm applied here are strongly correlated with the others. The presented model of
bacterial genome evolution, which was shown in the example of
B. burgdorferi
, should
give similar general results for other bacterial genomes because their DNA asymmetry
resembles that from the species analysed here.
Search WWH ::
Custom Search