It's going to be equal to the inverse probability, the probably of the data given theta times the probability of theta over the probability of the data.
We're not going to worry about the probability of the data in the bottom.
I haven't talked anything about the denominator, the probability of the data.
This file contains an informal estimate of the model's estimation of the probability of the data while training.
And if we knew something about the probability of the data and put it in the denominator, we can put it there.
So in a lot of what we're going to talk about and what you will do is when you're doing MCMC methods, the probability of the data, all it does is shift the surface up and down. in terms of it's just a constant, right?
So by using Metropolis-Hastings, using Metropolis, running it long enough that we've converged on the posterior distribution, they will ensure that we can write our equation in this way and disregard the probability of the data and that the relative density of our random samples in different intervals will reflect the true probability.
The notation BF01 is used to express the probability of the data given the null hypothesis (model 0) relative to the alternative hypothesis (model 1).
The prior probability is independent of the experimental MS/MS data, whereas the likelihood is the probability of the data, given the model.
The PMP is proportional to the marginal likelihood of the respective model, i.e. the probability of the data given the model, times the prior model probability.
In other words, the probability of the data item D i was assumed to be Pr D i = 1 i θ ∑ j = 1 N 1 j θ (29).
