The root motivation of this paper is to assist a novice user for graph drawing by generating different drawing for his data and let him choose the one which best fits his needs. We chose to use a genetic algorithm to generate these drawings. This paper focus on two main contributions which are ones of the first steps to achieve this goal. The first one consists in changing the way graph drawing GAs generally encode the genome. This allows a better abstraction between genotype and phenotype. In our project, each genome encodes a set of parameters for one vertex. Then a modified force-based genetic algorithm compute the resulting layout which is used to compute the fitness of each vertex genome. The second contribution is an enhancement of genetic algorithms (GAs) and cased-based genetic algorithms to make genomes more reusable. This evolution allows for an important increase in speed after a learning phase and also allows for an increase in the quality of the results and in the solution space coverage.