Models of nucleotide substitution

Models of nucleotide substitution are mathematical equations built to predict the probability (or proportion) of nucleotide change expected in a sequence.

Additional recommended knowledge

Don't let static charges disrupt your weighing accuracy

How to ensure accurate weighing results every day?

Essential Laboratory Skills Guide

Jukes and Cantor's one-parameter model

JC69 is the simplest of the models of nucleotide substitution.^[1] The model assumes that all nucleotides has the same rate (μ) of change to any other nucleotides. The probability that any nucleotide x remains the same from time 0 to time 1 is;

P_xx(1) = 1 − 3μ

P_xx(t) must be read; probability (or proportion, in this case it is equivalent) that x becomes x at time t. For the probability that any nucleotide x changes to any other nucleotide y we write P_xy(t). The probability for time t + 1 is;

P_{xx(t + 1)} = (1 − 3μ)P_xx(t) + μ(1 − P_xx(t))

The second part of the equation denotes the probability that the nucleotide was changed from time 0 and 1, but then got back to its initial states on time 2. The model can be rewritten in a differential equation with the solution;

Or if we want to know the probability of nuleotide x to change to nucleotide y;

With time, the probability will approach 0.25 (25%).

Kimura's two-parameters model

Mostly known under the name K80, this model was developed by Kimura in 1980 as it became clear that all nucleotides substitutions weren't occurring at an equal rate. Most often, transitions (changes between A and G or C and T) are more common than transversions.^[2]

Further Reading

• Yang, Z. (2006). Computational Molecular Evolution. Oxford University Press. 

References