SubstitutionModels.jl

What are subsitution models?

Substitution models are phenomenological descriptions of the evolution of genetic code. These models are Markovian, and do not explicitly depict mechanisms such as mutation or natural selection. Rather, they describe the relative rates of different changes, and those rates are assumed the capture the action of those mechanisms.

For example, when examining aligned sequences, a high number of transition substitutions are observed vs. the number of transversion substitutions. The mechanisms that cause this differential include mutation biases, and the action of purifying selection. However, the Kimura (K80) substitution model merely attempts to capture the effect of both of those mechanisms, rather than modelling the mechanisms themselves, by using a parameter that reflects the relative rate of transitions to transversions.

Models describing the substitution rates of nucleic acids, and amino acids both exist. Currently only nucleic acid substitution models are implemented in SubstitutionModels.jl.

Representation of models in code: