Where does phenotype collide?

The set of observable characteristics of an organism, its form, physiology, and behaviour, that arise from the interaction of its inherited genotype with the environment in which it develops. Wilhelm Johannsen's 1911 The Genotype Conception of Heredity used the paired terms, drawing a sharp line between the hereditary constitution and the visible traits it helps produce, and modern biology studies the genotype-phenotype map and phenotypic plasticity, the way one genotype yields different traits under different conditions. The referent here is the whole visible organism read against its genes and its surroundings, not a gene's reach outside the body, a disease pattern in one patient, or a decoded string in a search program.[1]

Unit

One organism's observable trait set, scored trait by trait

Recovery means

Not applicable

Measured by

Assessed by observing and quantifying morphology, physiology, and behaviour, often across environments to reveal plasticity

The idea that a gene's observable effects are not confined to the body of the organism that carries it but reach outward into artefacts and even into other organisms it influences, so that a beaver's dam or a parasite's manipulation of its host counts as part of the gene's phenotype. Richard Dawkins named the concept in his 1978 paper Replicator Selection and the Extended Phenotype and developed it at length soon after, arguing that natural selection acts on genes through every effect they have on the world. This is a distinct theoretical claim about where a phenotype ends, and it should not be read as the ordinary traits of a single body, a clinical picture of illness, or a candidate solution inside a computer.[2]

Unit

A gene's phenotypic effect, wherever in the world it lands

Recovery means

Not applicable

Measured by

Inferred by tracing a gene's influence on built structures, host behaviour, or other organisms, and testing its effect on the gene's own survival

The observable expression of a disease in a patient, the collection of signs, symptoms, and measurable features that mark how a condition presents, used to group patients who may share an underlying mechanism or differ in it. Peter Robinson's 2012 Deep Phenotyping for Precision Medicine argued for recording these observable features in fine, structured detail so that patients can be stratified and matched to the right genetic and molecular explanation. Here the phenotype is a disease as it shows in one person, read against a hidden genotype, and it is not the full trait set of a healthy organism, a gene's reach beyond the body, or an encoded solution in a program.[3]

Unit

One patient's disease presentation, coded as observed features

Recovery means

Not applicable

Measured by

Captured by structured clinical observation, standardised feature vocabularies, imaging, and laboratory tests, then linked to genotype

In a genetic algorithm the phenotype is the decoded candidate solution that an encoded genotype maps to, the actual object, a schedule, a route, or a design, on which fitness is then evaluated. Darrell Whitley's 1994 A Genetic Algorithm Tutorial laid out this genotype-to-phenotype mapping, in which a string of symbols, the genotype, is interpreted into a working solution, the phenotype, that the algorithm scores and selects. This meaning borrows the biological word by analogy but names a step in a search procedure, not a living organism's traits, a gene's effects in the world, or the way a disease appears in a patient.[4]

Unit

One decoded candidate solution, ready to be scored

Recovery means

Not applicable

Measured by

Obtained by decoding the genotype through a fixed mapping, then evaluated by the fitness function on the resulting solution