.png){kind=icon}
.png){kind=icon}
Return
DNA sequences design under many objective evolutionary algorithm
.png){kind=icon}
PRE
AI
DOI：10.1007/s10586-024-04675-1.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
DNA computing is conducted through reactions between DNA molecules, the quality of DNA sequences is directly influence on reactions. Following previous works, there are five metrics to estimate quality of DNA sequences and one constraint to follow. Evolutionary algorithms are widely applied in this field, conventional frames are often using multi-objective strategies to solve this problem. However, multi-objective strategies loss its efficiency in solving high dimensional problems especially Pareto Front is irregular. In this article, a many-objective evolutionary algorithm, R2HCAEMOA, is introduced to tackle with increased objective dimension. To increase diversity from beginning, chaotic mapping is applied to initialize decision variables of population. Since purpose of many-objective optimization algorithms is to find evenly distributed solution set on Pareto Front, decision makers are faced difficulty in solution selection. A method for choosing the most interesting solution from solution set is determined. Besides, an incremental scheme to generate a DNA sequence set is applied to enforce stability of evolutionary environment. The average values on each metrics are {0, 0, 56.00, 42.85, 0.15}, which the metrics are {continuity, hairpin, H-measure, similarity, variance of melting temperature}. Running time of our frame is significantly reduced compared with previous works. The results have shown our work is competitive among previous works and incline balanced value on each objective dimension.
Keywords:
DNA sequence design
DNA computing
Evolutionary algorithm
Many-objective optimization
Journal
C
IF:
4.1
Papers:
5.0K
Citations:
7.5K