Molecular evolution Flashcards
What is the Jukes Cantor model?
The simplest probabilistic model. The same substitution rate for all nucleotides, the rate for each nucleotide is 3*alpha per time unit. Does not consider the difference in amount of transitions and trans versions. Can be unrealistic since it assumes all nucleotide substitutions occur randomly. One parameter.
What is a probabilistic model?
When estimating sequence divergence in proportion to time, we expect to get a “molecular clock” that we can use in order to determine how diverged two sequences between for example two species are, based on how long ago they diverged. However, the number of substitutions are often underestimated because of hidden events, i.e multiple hits in the same position. Probabilistic models are used to correct sequence divergence to get more accurate estimations.
What is the Kimura Two-parameter model?
Invented as an improvement to the JC model. Takes into account that transitions occur more frequently than transversions. Rate of transitions = alpha. Rate of transversions = beta. Two parameters.
What is the General Time Reversible (GTR) model?
Six parameters, six different substitution rates = one for each substitution type. However, the reverse substitutions (eg. A-T and T-A) has the same rate.
What is the Unrestricted model?
Twelve parameters, one for each substitution type, including reverse ones. Very complicated.
What could be some factors that contribute to the inconsistency to the molecular clock?
The substitution rates vary between genes due to varying levels of functional constraint
Different organisms could have differences in the efficacy of the DNA repair mechanisms.
Generation time hypothesis: organisms with shorter generation times have higher mutation rates, because they replicate their genomes more frequently.
Timing of replication: late or early replication. Late replications have higher mutation rates.
There can be differences in selective pressures among different species.
Horizontal gene transfer or introgression can complicate the interpretation of the molecular clock.
If natural selection takes up much space in the molecular evolution, how can we then use the molecular clock that is mostly based on neutral evolution?
Fluctuating selection pressures is quite common among organisms due to fluctuating environments. This causes genes to evolve in a steady way that resembles neutral drift. Some genes are under positive selection, while others are mostly influenced by neutral mutations, therefore the rates differ depending on different regions in the genome. But the substation rates may still be similar in the homologous region between different species.
What is reinforcement?
The evolution of stronger pre-zygotic isolation by natural selection.
Choosing a partner from within population provides higher fitness –> leads to a selective advantage in alleles that increase the likelihood of mating within own population. The “discrimination” allele provides higher fitness than the “random mating” allele.
What are models of sequence evolution, why are they needed?
Models of sequence evolution are models used for estimating the rate of substitution among organisms and helps us to date different evolutionary events. The molecular clock hypothesis states that the substitution and mutation rate in DNA is relatively constant among different organisms, which means that a larger amount of accumulated differences between two species should be proportional to the time since their last MRCA, i.e since they last diverged. Different models have different assumptions and vary in the number of parameters, for example the Jukes-Cantor model is the simplest model since it only has one parameter. It assumes the substitution rate is the same for all types of nucleotide substitutions. These models are used for inferring evolutionary relationships between species by estimating branch lengths of a phylogenetic tree.
Which model should be fit to the data? Discuss approaches for motivating/supporting your choice?
Different types of data fit different types of models. Maximum Likelihood Estimation (MLE) is used to find the model with the highest fit for the given data. MLE calculates the likelihood of observing the data given the model and the parameters. Each model is based on a set of assumptions on how probable different types of nuceltide substitutions are. This calculation is performed for different models with the same data, and the model that provides the highest likelihood is used for the data. These probabilistic models are used for correcting the estimation of the number of substitutions that have occurred in the genomes of two species since their divergence. In theory, assuming the molecular clock is accurate, the number of substitutions should reflect how diverged the species are. However, the number of observed differences does not equal the number of substitutions that most likely have occurred, because of hidden events. Probabilistic models are applied to correct this number and achieve the most likely number of substitutions that have actually occurred.
What is the prediction of the Molecular Clock hypothesis? Explain some factors that may cause deviation from the molecular clock! Why does the substitution rate vary across the genome?
See key concepts 1.
