Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I have a Venn Diagram that looks like this:

$$A) \, 213 \quad B) \, 160 \quad A\cap B) \, 100$$

The items from $A$ come from a population where their probability to be selected is $\frac{313}{12800}$.

The items from $B$ come from a population where their probability to be selected is $\frac{260}{1407}$.

How can I calculate a null model that tells me the number of shared items expected by chance if I create a Venn diagram for that data?

share|improve this question
    
Presumably the $213$ items are $A$ and not $B$ and the $160$ are $B$ and not $A$. However normally you would have one population, some of which has characteristic $A$ and some of which has characteristic $B$. Then the population outside both $A$ and $B$ would be the rest of the universe, here everything except $473$. Is that $12327$ or $934$? –  Ross Millikan Feb 20 '13 at 18:36
    
The items (313 and 260) are expressed genes in a given condition. And the totals (12800 and 1407) are total sampled genes. What I want to prove is if this A & B is not a product of chance. –  biojl Feb 21 '13 at 8:18

1 Answer 1

up vote 0 down vote accepted

If $A$ and $B$ were independent, you would expect the same fraction of $B$'s to have $A$ as the whole population. Of the $260 B$'s, $100$ are also $A$'s, about $38\%$. But of the whole population, $A$'s are rare-a little less than $4\%$. Similarly the other way around. They are far from independent.

share|improve this answer
    
I would like to use a statistical test. Do you think a Fisher Test (observed vs expected) will suffice? –  biojl Feb 22 '13 at 14:37

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.