Page 11699 of this paper has one approach to what you want: basically, you compute the relative Gibbs free energies of the possible stable conformers, assume they can interconvert, and use a Boltzmann distribution to work out what the percentages are.
First, we selected a reference structure $R$ for a given species (neutral or protonated).
Next, for every structure $M$ other than the reference structure $R$ we determined the
equilibrium constant $K_M$ $$K_M = \frac{[M]}{[R]}$$ from the difference in Gibbs free energies for $M$ and $R$. The fraction of $M$ in the equilibrated sample is given by:
$$x_M = K_M/\left(1 + K_1 + K_2 + ...\right)$$
where the sum in the denominator goes through all structures for a
given species. The fraction of $R$ in the sample is:
$$x_R = 1/\left(1 + K_1 + K_2 + ...\right)$$
They don't include it in the paper, but the way to calculate an equilibrium constant $K$ from a difference in Gibbs free energy $\Delta G$ is:
$$ \Delta G = RT \ln K $$
where $R$ is the molar gas constant, and $T$ is the temperature (in Kelvin).
The tricky part is then working out what your stable conformers are, generally using some form of either guesswork or conformation sampling algorithm. They use a genetic algorithm of some sort in the paper I linked to, but given that methylcyclohexane isn't that complex a molecule, you might want to just take a look at an energy diagram and discussion of conformers of cyclohexane for some hints.
If you're getting deep into molecular dynamics, you could also use some sort of algorithm designed to use dynamic simulation to sample a wide area on a potential surface, like metadynamics or umbrella sampling, to find your conformers, and probably calculate your free energies at the same time.
You can use a static method to calculate the free energies instead, though -- the method you use to do this will depend on what sort of accuracy you're after, and may vary from a classical forcefield technique through Density Functional Theory to coupled-cluster theory (in ascending order of computational expense).