Fede
Posts:
14
Registered:
6/8/10
|
|
best test to determine if two proportions are different
Posted:
Aug 21, 2012 12:08 PM
|
|
Dear all,
Let me first state my problem. I am analyzing a genome and its genes.
I can calculate proportions of different types of genes in the genome
(pseudogenes, lincRNAs, coding_genes...). Now, I have particular
genomic regions of interest and the corresponding list of genes
belonging to these regions. I want to test if genes of type X (e.g.
pseudogenes) have the same frequency in this region as in the whole
genome.
I have been playing with different tests. Firstly, the binom.test,
assuming that p=freq_in_genome, n=number of genes in the regions of
interest and x=number of successes (number of genes of type X).
However, I am not sure if this is the most appropriate test and have
been looking for alternatives to test for equal proportions: fisher's
exact test, on one-sample z-test, prop.test in R, etc... (some of them
are probably the same ¿?)
Which one is the most appropriate test?
Many thanks!
Federico
|
|
