MACS file/gene list

[araposo]

Hello.
When I first read about your tool, it looked exactly like what I was
looking for: a nice, easy and quantitative way to integrate ChIP-Seq
wit expression data.
Unfortunately, I'm not being able to make it work for me... I was
wondering if could answer any of these questions:

1. I have uploaded a MACS type file, selected distance to TSS (-100kb,
+100kb, mm9) and then copy-pasted a list of 2272 genes into the
microarray window (selecting 'gene symbol' as the gene list to
consider).

Now, I know from other tools (PeakAnalyzer) that these binding events
give me an annotation of 15433 genes in these conditions, of which 800
overlap with my expression data.

So, my problem is that, ChIP-Array does seem to produce the annotation
for BE but then it doesn't find any overlap between the lists. This is
the output:

Enriched P-value 0.000E+00
#Total Gene in Genome 21764
#Total Target Gene 15435
#Total Differentiated Expresstion Gene 2242
#Differentially Expressed Target 0

My question is, am I doing something wrong or is this a bug?

2. This other question relates to the size of the universe considered.
Binding events are virtually infinite, but the mm9 annotation predicts
~37k genes. However, a microarray chip may contain much less genes
represented; in my case, 28853. In addition, I see you have considered
21k for the genome. How do you calculate the p-value for the overlap?

Cheers,
Alexandre

[Jing]

Hi, Alexandre.
I guess the problem is caused by the large Target Gene set, which is
beyond the limit of the web server.
Please try a shorter distance to TSS for the peak annotation.

For the question 2 in your case, the N is 21764, m is 15435, n is 2242
and k is the overlap of these two sets for the formula in the
following link:
http://en.wikipedia.org/wiki/Hypergeometric_distribution

Best,
Jing