1
$\begingroup$

I would like to compute co-occurence and mutual exclusivity for a set of events divided in two groups (A and B). Each event E_{x,y} represents if a mutation occurs in the gene x from sample y These two groups of events are encoded in two matrix (A and B) with the same number of columns and rows ( columns = samples ; rows = genes). Cell values can either by 1 (mutation occurs) or 0 (no mutation occurs). A_{x,y} represent an event of type A for gene x from sample y. B_{k,l} represents an event of type B for gene k and l.

I would like to test co-occurence or mutual exclusivity for each pair of genes. Thus comparing each pair of A_{i,.} ; B_{i,.} where i is 1..number of genes.

Is a fisher exact test a good option ?

Thank you

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

Yes, Fisher exact test seems a reasonable option to me. For each gene you have a 2x2 table with for numbers of occurences (A and B, A and not B, not A and B, not A and not B). The odds ratio (together with the p-value) will tell you whether you have co-occurences or mutual exclusivity.

Improve this answer
$\endgroup$
3
  • $\begingroup$ Ok thank you. And is fisher test ok if there is might be a causality effect between events from group A to B. So to check if event A_{x,.} is = 1 is there some event B more prone to be 1 ( or 0 ) $\endgroup$
    – Nicolas Rosewick
    Mar 10, 2019 at 8:52
  • $\begingroup$ No, the the Fisher test does not tell you anything about causality. It can be that A causes B, that B causes A or that another cause C causes both A and B. $\endgroup$
    – citronrose
    Mar 11, 2019 at 11:27
  • $\begingroup$ ok thank you. Reading a little bit about testing causality and indeed it's not trivial .. $\endgroup$
    – Nicolas Rosewick
    Mar 11, 2019 at 12:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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