genes

Terms from Artificial Intelligence: humans at the heart of algorithms

Genes are areas of a chromosome that form a coding unit, raher like a word or sentance in text. At a low-level genes are related to the production of various proteins, but at a higher level give rise to inherited traits such as hair colour. In humans, other animals and many plants, genes come in pairs one inherited from each parent, but there are exceptions, for example, those missing on the (rather stunted) male Y chromosome, or occasionally addtional chromosomes. The pattern of genes in an animal or plant is called the genotype whereas the way this appaers (say eye color) is called the phenotype. Without {DNA}} testing only the phenotype is observabe and able to make a difference to the ppearance, behaviour and abilities of the creature or plant.

.Genetic algorithms seek to mimic the way genes are inherited from parents and the survival of prodgeny is dependent on derived features (phenotype).

In early genetics, genes were regarded as having isolated and rule-like behaviour: a blue-eye gene and a brown-eye gene (genotype) giving rise to a brown eyed person (phenotype). However, it is now known that the mapoing is far more coplex, with many genes and environmental factors giving rise to the observable phenotype, and also complex interactions within the genome: genes turned on and off, and genes that interact stringly often clusterd togetger on the chromosome so thet they are usually inherited as a unit.

This more nuanced understanding of biological genes suggests potential insights for genetic algorithms, although the majority still are based on 19th century genetics.!

Used on pages 76, 77, 164, 191