Here is a current problem from my research project I could use some help with.

Of four colonies from genetic background A, two have phenotype X.

How many colonies from genetic background B would I have to test to see whether the frequency of phenotype X is significantly different from 2 of 4?