EvolveJS is a lightweight genetics library modeled on Mendelian inheritance and Darwinian selection. When implemented into any self-replicating JavaScript object instance, the library can be used to produce descendant instances whose quantifiable variable
npm install @matthewmain/evolve.jsEvolveJS is a lightweight genetics library modeled on Mendelian inheritance and Darwinian selection. When implemented into any self-replicating JavaScript object instance, the library can be used to produce descendant instances whose quantifiable variables are handled as heritable genes. Sequential replication iterations are then subject to the laws of Mendelian inheritance, gene mutation, and, when Darwinian environmental selection pressures are applied, in time, evolution.
This guide will walk you through how to apply EvolveJS to JavaScript objects that represent biological organisms. Using a species of flowering plants as an example, we'll go from creating first-generation organism objects to producing new, unique offspring objects that have inherited a combination of their parents' genes. All of the code from the examples has been implemented in _Kiss The Sky_, an indie video game based on plant evolution. For reference, check out the _Kiss The Sky_ source code here on GitHub, or have a look at a live application alongside the code here on CodePen.
First, you'll need to add the EvolveJS library to your project and reference the source in your html.
``
Or, you can reference the CDN directly.
`
EvolveJS is published as an npm package, so you also have the option to include it in your project as a module, if you prefer. Just install the package in your project's root directory using npm i @matthewmain/evolve.js. Then all you need to do is require the library in your source files using the following naming convention.
`
var EV = require("@matthewmain/evolve.js");
A genome is a genetic blueprint for a generic species body. So, for our example, the first thing we'll need to do is create a genome for a plant species. EvolveJS can handle any number of distinct species genomes at once, but for this guide we'll create just one. We'll use the EV.addGenome() method. It takes two arguments: 1) _the new name you want to give the new species_, and 2) _the species's reproduction type_. A species can have an "asexual", "sexual", or "autogamous" reproduction type. Asexual species spawn offspring with copies of just their own genes. Sexual species generate offspring by combining genes from a male/female pair. Autogamous species, like many plants, are capable both of reproducing with other organisms and self-fertilizing. In the spirit of _Kiss the Sky_, we'll call our plant species a "sky plant". Because it will be a flowering plant that can both cross-pollinate and self-pollinate, we'll give it an autogamous reproduction type.
`
var skyPlantGenome = EV.addGenome( "skyPlant", "autogamous" );
Now we'll need to add some genes to the plant genome. A gene represents a feature (like eye color or height). It is composed of two alleles (each of which represents a specific trait, like brown eyes or six feet tall). If a gene's alleles are different, one will always be more dominant in relation to the other. The dominant allele's trait will be expressed in the organism, either in full or in part. The recessive allele's trait will only be expressed in part or not at all. For this step we'll use the EV.addGene() method, which takes the following nine arguments.
1. _Species genome_ (species genome object). The species genome you're adding the gene to. This can be either a variable assigned using EV.addGenome() (as in the example above) or an EV.species. object that the method automatically generates."complete"
2. _Gene name_ (string). The new name you'd like to give the gene.
3. _Dominance type_ (, "partial", or "co"). "Complete" means only the most dominant allele's value will be expressed and the recessive allele will be ignored. "Partial" means an average of the two allele values will be expressed. "Co" means that both allele values will be expressed in different locations or in different ways, such as the colors of patchy fur. (Note that co-dominant values are be expressed as an array of two allele values.)"count"
4. _Expression type_ ( or "scale"). Use "count" for genes whose values need to be expressed exclusively as integers (like leg count or leaf count). This will ensure that any averaged or mutated values are always returned as whole number values. Use "scale" for genes whose values can represent any point on a sliding scale (like height or speed) and can be handled as decimals.null
5. _Initial value_ (number). The gene's initial value as expressed in first-generation organisms and carried by a pair of identical alleles. Initial values can be integers or decimals.
6. _Mutation frequency_ (integer). How often the gene mutates, as average reproduction events per mutation. Higher is less frequent.
7. _Mutation range_ (number). The range in which a new mutated value can occur, with the current value at the middle of the range. For example, if the allele's current value is 10 and you want new mutations to range from 8 to 12, then the mutation range should be 4.
8. _Minimum value_ (number or ). The minimum value that any trait can have. Mutated allele values cannot fall below it. An argument of null allows values to mutate to increasingly lower values without limit. null
9. _Maximum value_ (number or ). The maximum value that any trait can have. Mutated allele values cannot rise above it. An argument of null allows values to mutate to increasingly higher values without limit.
EvolveJS allows you to add an unlimited number of genes to a genome and create an extremely complex organism. But here we'll include just a few basic plant characteristics as example genes.
` `
EV.addGene( skyPlantGenome, "stalkStrength", "partial", "scale", 0.75, 5, 0.1, 0.7, 0.8 );
EV.addGene( skyPlantGenome, "maxLeafLength", "partial", "scale", 5.5, 5, 3, 4, null );
EV.addGene( skyPlantGenome, "flowerHue", "complete", "count", 130, 5, 50, 0, 260 );
A genotype is a genetic blueprint for a specific organism's body. It is composed of all of that organism's genes, each of which includes specific allele pairs. The organisms that will compose our first generation of plants will need to be assigned their own genotypes. (All of the organisms in later generations will automatically inherit a combination of their parents' genes to form their own new genotypes.) EvolveJS offers a couple of options for generating new first-generation genotypes. You can use EV.newStandardFirstGenGenotype() to create any number of genetically identical first-generation organisms, or you can use EV.newRandomizedFirstGenGenotype() to create a variety of unique organisms whose genes are slightly different. Both take one argument, a _species genome_ as a species genome object. We want to start with a diverse population of sky plants, so we'll use randomized sky plant genotypes.
A phenotype represents the organism's physical body. It is composed of all of the organism's expressed traits, each of which is determined by the rules of dominant and recessive gene expression. EvolveJS will automatically generate an organism's phenotype with EV.generatePhenotype(). It takes one argument, the organism's _genotype_ as a genotype object.
EvolveJS can be applied to any Javascript object constructor that represents an organism. For our example, let's use a simple plant object constructor. We'll give it a genotype parameter so that later generations' new genotypes can be passed into the constructor. We'll also give it a genotype variable, a phenotype variable, and gene variables for each of the genes we created above. The gene variables need to be assigned their phenotype value, which is indexed in the phenotype object as the gene's name with an appended "Value" substring. For example, the "stalkStrength" variable will be assigned as phenotype.stalkStrengthValue.
`
function Plant( genotype ) {
this.genotype = genotype;
this.phenotype = EV.generatePhenotype( this.genotype );
this.stalkStrength = this.phenotype.stalkStrengthValue;
this.maxLeafLength = this.phenotype.maxLeafLengthValue;
this.flowerHue = this.phenotype.flowerHueValue;
}
From here, we can create any number of sky plant organism objects. Let's create twenty new sky plants with unique genotypes and store them in an array as an initial first-generation population.
`
var skyPlants = [];
for ( var i=0; i<20; i++ ) {
var newSkyPlantGenotype = EV.newRandomizedFirstGenGenotype( skyPlantGenome );
skyPlants.push( new Plant( newSkyPlantGenotype ) );
}
Now that we have an initial population of sky plants outfitted with EvolveJS, each plant is now equipped to reproduce and generate new offspring. Whenever your program is ready for organism reproduction, from here all you need is the EV.meiosos() method to generate new offspring organisms. It takes three arguments, 1) the _species genome_, 2) the _first parent organism's genotype_, and 3) the _second parent organism's genotype_ (a second parent genotype is optional for autogamous species, and should be omitted for asexual species). Because sky plants are autogamous, whenever they self-pollinate we can either include just one parent genotype, or include the same parent genotype twice as the first and second parent genotype. Whenever sky plants cross-pollinate, we'll need to include both parent genotypes.
(Note that for sexually reproducing species, you will need to provide one male and one female genotype as parent genotypes. An organism's sex is stored under the namespace. Sexually reproducing organism sexes are assigned randomly as either "male" or "female", an asexual organism's sex is always "none", and an autogamous organism's sex is always "hermaphrodite.")
For our example, we'll create a new offspring sky plant genotype with two parents who have cross-pollinated. We'll then pass the genotype object into the skyPlant organism object constructor and add the new sky plant to the population array.
`
var childSkyPlantGenotype = EV.meiosis( skyPlantGenome, skyPlants[0].genotype, skyPlants[1].genotype );
skyPlants.push( new Plant( childSkyPlantGenotype ) );
Every time your program reproduces organisms using meiosis events, each new offspring object will automatically have a genotype constructed from its parent or parents' genotypes and a resulting phenotype based on the laws of Mendelian genetics. Some genes will mutate during reproduction, and the offspring carrying these genes will slightly alter the species genome. If Darwinian selection pressures are applied and the species reproduces for many generations, the species genome will gradually evolve.