Identification and Exploitation of Linkage by Means of Alternative Splicing

Alternative splicing is an important cellular process that allows the expression of a large number of unique cell-specific proteins from the same underlying strand of DNA, and thereby drastically increases the organism’s phenotypic plasticity. Its emergence is facilitated by the modular composition of genes into numerous semi-autonomous building blocks. In artificial evolution, such modular composition is usually unknown initially, but once learned may greatly increase the algorithm’s efficiency.
In this paper, an abstract interpretation of alternative splicing is presented that emulates some of the properties of its natural counterpart. Two appoaches, both based upon a simple (1+1) evolutionary algorithm, are described and shown to work well on established benchmark problems. The first algorithm, eAS, is designed for cyclical dynamic optimisation problems: it systematically merges the problem variables into groups that capture the properties exhibited by a finite number of successive states and reuses that information when required. The second algorithm, iAS, employs a systematic search to identify a sub-set of variables for which simultaneous inversion affords an increase in fitness. This approach seems particularly useful for problems that have many local optima that are far apart in the search space. Results from a systematic series of experiments highlight the intrinsic attributes of each algorithm, and allow analysis in terms of the identification and exploitation of linkage.