The results showed low expression of A and overexpression of B in Tumor tissues.
We expect them to show a negative correlation, but they show a significant positive correlation. Can you explain this situation?
Can we say the more A decreases, the less B increases?
Have you ever seen such a situation?
Thank you
How did you plot your correlation graphs? By fold change against the regulatory genes or compared against a control sample?
Is your "low expression of A" and "overexpression of B" compared against the same thing?
You mean a situation where your beautiful theoretical idea turns out flat wrong, once you test it experimentally? Yeah. That's called science!
John Schloendorn I don't think this is a question about whether the hypothesis is correct or not. It's a question about correlation statistics. (The poster should correct me if I'm wrong.) Correlation is either positive or negative. A positive correlation graph means that as x goes up so does y. A negative one is where as x goes up, y goes down (and vice versa).
Yuan-Yeu Yau
No, I did't see it. I just saw "low expression of A" and "overexpression of B". But sample with low expression of A showed overexpression of B.
Shen-An Hwang I used from 2(^-d Ct) & 2(^-dd Ct). Both gave the same answer.
my samples were tumor tissues and their adjacent tissues. To obtain dd Ct, I used the following formula (GeneX software):
dd Ct= d Ct (tumor) - d Ct (mean of adjacent tissues)
Shen-An Hwang In fact, I want to know if this is possible or I have used the wrong formula.
Marjan Ostovar Pour I was wondering how you did the correlation analysis. So when you input the ddCt from the two genes for all your samples, the Pearson correlation (r) is a positive value?
So it is possible that it can be a positive correlation because as your value of A goes up so does B, even though A is always lower than B.
Shen-An Hwang
Yes. Actually I used from spearman cor, because my data was non-parametric.
So, do you think is that possible? Have you ever seen this in articles?
Marjan Ostovar Pour This was what I figured, that even though your A is expressed at lower levels compared to B, when A goes up so does B. So the expression of A and B goes in sync. So, yes, this is possible. It's been a while since I've looked at expression data, so I can't even tell you when I would have seen a similar result.
The r value is not strong, so the correlation is only so so.
The way to interpret the data isn't that low expression of A leads to high expression of B. The correct interpretation should be that compared to non-tumor tissue, tumors have decreased expression of A and increased expression of B. And that there is a weak positive correlation between expression of A and B.
Shen-An Hwang Thank you very much. So I can say that my way is not wrong, but My Supervisor is not convinced by these explanations.
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics
- Gene Expression