Entering edit mode

7.3 years ago

Ben
▴
60

I want to analyze RNA editing with RNA-seq data, but how to calculate the p value if I detected one RNA editing, A-to-I. binom.test() might be ok. However, I do not know how to set the p value of bionm.test(). the default p value argument is 0.5, Could anyone help me ?

It seems you are confused about what a p-value is. A p-value is a measure of statistical significance in the context of null hypothesis testing. In this context it is the probability of the test statistics being more extreme than what is observed when the null hypothesis is true.

The binomial distribution with parameters n, p is the distribution of the number of successes in a sequence of n independent trials each with probability p of success.

So, the argument p to binom.test() in R can not be not a p-value. It is the expected probability of success in a binomial distribution. To do a binomial test, you need to define n, the number of independent trials and p the a priori probability of seeing an editing event. I don't know much about RNA editing but I would be surprised that detection could be reduced to a simple binomial test. I am also quite sure that there are papers on the subject. Have you read them ?

Thanks for your reply! Yes, you are right! The argument p in the binom.test() is the priori probability. My hypothesis is that one base might be mutated or not. So I want to use binom.test() to analyze significant RNA editing changes.

Please use the 'add reply' button to reply to a comment, do not create an answer. This keeps the discussion organized and avoid misleading users into thinking a question is answered when it is not.

I have some doubts about the validity of your approach. For one, the binomial test assumes the trials to be independent. This is clearly not the case with nucleotide sequences and also for RNA editing because editing events are thought to be clustered. I would suggest you talk to someone more familiar with the domain.