This lesson is going to be about mutations. There's a lot of detail about mutations, some of which this lesson is going to explore. For our purposes, mutation is a source of variation. Once that variation exists, it's shaped by selection, by drift, and later by migration. But mutation itself has a very small effect on existing variation. Its main contribution is just to add variation. As such, the mechanics of mutation are the mechanics of adding variation in genetics, but the fate of mutants is going to depend largely on selection and drift. And so to that end, the most important dichotomy here is whether a mutation has an effect on fitness (non-silent) or no effect on fitness (silent).
The other important aspect is that, even if it lacks a direct effect on fitness, a mutation can change some other property of transmission. In the case of an inversion, for instance, individuals with an inversion will have different recombination chances than individuals without it. This may have no direct effect on their fitness, the covariance of their traits and reproductive success, but it may have a big effect on the transmission bias that results from it, because they can't split loci within the inversion with other organisms. Recombination is distinct from mutation. It's a distinct process, but it is also a form of adding variation, so we're kind of lumping it together. For both the mutation and recombination parts of this, the important thing is to think about being a locus or a sequence, and what your fate is.
A single nucleotide polymorphism, a SNP, is what we're going to look at here first. SNPs are useful because, no matter what they do, whether they have a fitness effect that's large, small, or zero, they are not changeable by recombination. They are a single nucleotide. They are the smallest possible unit for a locus, and as such, a SNP itself cannot be recombined. A SNP can be recombined into a new combination with other SNPs, but if you think of a locus as a gene, I could have a gene with 2 mutations in it, and then recombination could produce a new gene that has only one mutation. A SNP is not that way. It is atomic, in the indivisible sense: a heritable variant.
So for this first part, there is an actual gene from a bacterium shown below, along with a codon chart. Here, what you're going to do is click on a single base pair and change it to a different base pair, delete it, or insert a new letter in front of it. Once you've made your mutation, you get to decide what it does. If it is non-silent, you can affect the s and the h value for your new allele, and then you can affect the population size that it's going to evolve through. If it is a silent mutation, a synonymous change, you can only affect the population size. Either way, once you've made your new mutant and set things accordingly, you can hit Go and watch 1,000 populations, each with your mutant arising in a single individual, evolve over time. And you can look at in what fraction of those populations your mutation went extinct, in what fraction it's still present, and in what fraction it went to fixation. You'll do this 8 times for 8 different mutations, and then look at the summary results for all of them.
SNPs are convenient because they can't be recombined. Any given SNP has exactly one parent. You have 2 copies of each locus, each base pair, in your body. Each of them came from one of your 2 parents, minus your own personal mutations. Each of them came from one of your 4 grandparents. Each of them came from one of your 8 great-grandparents, such that each individual nucleotide has its own ancestry, and those ancestries can differ between nucleotides. One nucleotide you may have inherited from your father's father, and the next nucleotide beside it you may have inherited from your father's mother. Each one has a single ancestral source and can be tracked separately. Your genome, on the other hand, and any chunk of DNA larger than one base pair, can have multiple ancestries due to recombination.
Recombination is what sexually reproducing organisms use as another means of generating variation. Just like mutation, the cost of it is reduced heredity. You are creating new variants in copies of yourself that are not perfect copies, and so are less like you than they could be. You are not passing it on perfectly. You are not perfectly copying.
What you're going to do in this is take the role of a sequence. The first panel will show a population of sequences; yours is red. You can click on any other sequence to see what it looks like. In your target panel, you will choose each generation whether you want to copy yourself, that is, reproduce perfectly, or mate, that is, recombine. You can go one generation at a time, or jump by 5 generations. I strongly recommend using the practice toggle heavily in this one. Your goal is to try and reach high frequency in the population. What you should see is that there are multiple ways to do this. You can have many genomes that are exactly you, or you can have a lot more genomes that have some pieces of you, where you, here again, are the red sequence of nucleotides. The panel below shows, over time, many populations carrying out your same strategy, with a drop-down menu letting you see alternative randomization histories, given there's a stochastic element to this.
Again, you're going to be a particular sequence of nucleotides in this scenario. Here, we're going to run many populations, and, unlike in part B, we're going to add mutation back in. So this activity involves recombination and mutation. Again, you're going to choose whether or not to involve yourself in recombination by selecting whether you want to mate or copy. Unlike before, you're going to set all your generations at the start. You're going to have a preset strategy. You can choose from a drop-down menu, or you can manually click each generation for what you want to do. But you set all your generations' worth of strategies, hit Go, and you'll observe how your sequence, red, expands or contracts in each population. There will be 100 populations; 10 of them are shown to you.
The big difference here is that mutations are going to occur that are going to affect some trait. What value of that trait is optimal may change over time, or it may stay steady. You will have the option to toggle that in the target menu. New mutations are, generally speaking, going to have a bad effect. How bad that effect is, you can also control in the drop-down menu. Likewise the number of generations you're going to run: you will also control that in the drop-down menu.
Below your population panel summary on the left, you'll also see one population in detail. This will show the genomes of individuals in there, including red boxes for loci that are inherited from you, gray and black boxes for loci from other individuals, and, importantly, little black dots that are the locations of individual mutations. Below that is a drop-down menu where you can see how many copies of you there are (where you, again, are the sequence) over the generations in your populations; how many whole chunks of you, how many unrecombined chunks of you, there are over time; the trait values associated with each genome, along with the value that selection was targeting each generation; of the genomes that have some piece of you in them, the fewest number of harmful mutations; along with the number of genomes alive, as these populations can, to a certain extent, crash and die. If the number of bad alleles piles up and a population's fitness drops below replacement, the population crashes.
This one will be similar to part C, with the added complication that we can now add an inverted portion to the chromosome. That is, in part C, you are an arbitrary sequence. Now, you're basically a set of loci: an inversion. That inversion will block recombination. That is, being an inversion when you're rare means that you have to reproduce by pure copying, whereas when you're common, you may end up in an individual homozygous for the inversion, in which case recombination will work as normal. What you're going to control is whether you will have an inversion or not in your target, and whether or not 2 copies of your inversion are both good, being homozygous at high fitness, are the same as heterozygous for the inversion, or are worse than heterozygous for the inversion. You will then run and try to hit the targets for the frequency of your haplotype.
Haplotype refers to the type of haploid, and it's going to be a generic word we're going to use a lot from here on forward, describing basically a chunk of DNA inherited together. An inversion is an extreme case of this, where the chunk of DNA has to be inherited together until it reaches high frequency. But a haplotype can be any chunk of DNA inherited together. It's going to be a word we're going to see a lot in migration, and it's a useful thing to track in general.