Project funded by AIF / ProInno
The program computes optimised oligonucleotides in such a way that these are optimized for a following gene synthesis. For the synthesis a single step PCR procedure was assumed, as it is schematically represented in the following picture.
Picture 1: Single-Step PCR according to Stemmer [1] (schematic)
The black and grey bars symbolize the primers of the top and bottom strand. During the PCR-cycles ever more oligonucleotides hybridize, so that larger DNA pieces develop. The assembled genes can be further amplified in a second PCR reaction.
Using a former developed Thermodynamics-module [2-6] the software computes the oligos in such a way that during the gene synthesis the following applies:
This can be achieved by diversifying the overlap regions and simultanious introduction of silent mutations[7]. The problem is quite complex, because a variation of the overlap concerns the melting temperature of the adjacent ranges, as well as the energetic characteristics of possible foldings.
On the other hand a permitted codon-swap (e.g. TTT->TTC) leads to the change of the codo usage, to the change of the melting point in the range concerned, as well as to the change of the hairpin probability for up to three Oligos. A DNA sequence of 1000bp and a hypothetical oligo-legth of 20bp leads to already 3^400 possibilities (configurations) if one assumes three variations of the overlap borders and three allowed silent mutations per codon.
Solution - Evolutionary Algorithm:
As optimization strategy an evolutionary Algorithm[8-11] was developed. This consists of the following elements:
The temporal operational sequence of such an optimization protocol presents itself for example as follows:
The so-called evolution window is of special importance. This is the range of meaningful mutations, which may not fail too strongly and not too weakly, so that an improvement of the Fitness arises. The so-called mutation-increment must be selected correctly.
With tournament selection a specified group of individuals (usually 2) are chosen at random from the population. The one with the better fitness is selected to be the survivor.If more than 2 are used, the process is to select the best from the set. The selection pressure arises as a result of the number of mutants competing in each case[12].
Picture 2: Principle of the Tournament Selection. The population size remains constant.
The following screenshots and diagrams refer to optimized oligo-variants for the synthesis of Methyltransferase_gidB. The Codonusage was optimised for abies grandis. The reaction parameter temperature is defined within close borders. Potential secondary structures were eliminated. The codonusage was optimized. All features were optimised simultaneously.
Picture 3: Dialog for parameter input. Batch processing for optimisation of the optimisation algorithm parameters.
Picture 4: Distribution of melting points before (blue) and after (red) optimisation.
Picture 5: Optimisation of the codon usage.
Picture 6: Elimination of energetically favourable hairpin configurations (red bars).
Picture 7: the Oligo-Browser offers the possibility of additional selection.
Picture 8: Evolution of the total mean fitness (quality of oligos) in tournament mode.
References:
[home]