Question: Advice for gene overlap with direction between two sets.
0
gravatar for simplitia
15 months ago by
simplitia40
simplitia40 wrote:

Hi so I have a scenario where I have 200 samples. Each samples I test for 10000 genes. I have two events, call it A and B. Whereby for each sample there can be x genes with gene.up and gene.down. What I want to do is compare if events A and B are similar: I want to see if the intersection is significant. To visualize this I do a venn diagram and see if the intersection is significant. Normally I will do a fisher exact test or hypergeometric test. However its strange here because I have to account for sample, direction and gene. So its coded like this. Sample1.gene.up only this will be consider a match. My question is what then is the total population. For example, if total gene was 1000 is the total population then, 1000 * 2 * n samples. The two because gene can be up or down. Finally it would look something like this. I'm using R.

q = length ( intersect ) 
m= length( n1 )
k= length(n2)
n= 1000 * 2 * total.sample - m


phyper(q,m,n,k,lower.tail=F)

for a fisher test it would look something like this.

total.sample = 200
m =matrix ( c(
    1000 * 2 * total.sample 
    , 400
    , 500
    , 700
)
,nrow=2)

fisher.test ( m , alternative = "greater")

I need advice if I'm doing this correctly? especially if the total population is is correctly calculated? thanks!

statistics chisquare R • 357 views
ADD COMMENTlink written 15 months ago by simplitia40
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1660 users visited in the last hour