Evolutionary Studies - Gene Expression Programming - page 444

selection pressures to get rid of imperfect genotype/phenotype mappings.
The fact that the noncoding regions of GEP allow the creation of a perfect
genotype/phenotype mapping, further reinforces the importance of neutral-
ity in evolution, as a good mapping is essential for the crossing of the pheno-
type threshold.
On the other hand, the reason why neutral motifs within structures, be
they parse trees or proteins, can boost evolution, is not so easy to understand,
although I think this is another manifestation of the same phenomenon of
recombining and testing smaller building blocks. In this case, the building
blocks are not entire genes with clear boundaries, but smaller domains within
genes. Indeed, in nature, most proteins have numerous variants in which
different amino acid substitutions took place. These amino acid substitutions
occur mostly outside the crucial domains of proteins such as the active sites
of enzymes and, therefore, the protein variants work equally well or show
slight differences in functionality. At the molecular level, these variants con-
stitute the real genetic diversity, that is, the raw material of evolution. The
neutral motifs of gene expression programming play exactly the same func-
tion, allowing the recombination and testing of different building blocks and,
at the same time, allowing the creation of neutral variants that can ultimately
diverge and give birth to better adapted structures.
12.5 The Higher Organization of Multigenic Systems
Gene expression programming is the only evolutionary algorithm that deals
with genes as separated entities, tied up, however, in a more complex struc-
ture - the chromosome. From the analysis of the previous section, it is clear
that multigenic systems are far superior than unigenic ones. Here, we are
going to make a more systematic analysis by comparing multigenic and
unigenic systems with exactly the same chromosome length. The problems
chosen for this analysis are exactly the same of the previous section (that is,
the same function finding problem and the same sequence induction prob-
lem) using also the same general settings (see Tables 12.4 and 12.5).
For this analysis, a common chromosome size of 75 was chosen for three
different chromosomal organizations: a unigenic system with h = 37; three-
genic chromosomes with h = 12; and five-genic chromosomes with h = 7.
The performance of all these systems was measured in terms of success rate
and is shown in Tables 12.4 and 12.5.