Hello everyone
I would like to ask question about situations where gene expression values are used to correlate with clinical features of samples under investigation. For example, in an experiment, it is observed that lower the CT value of gene A gets, higher the value of biochemical parameter goes lets say glucose levels and correlation value was therefore found to be r= -0.67, P
Hi Azeem,
I agree with the others above and also suggest you look at "The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments" for assistance on qPCR. The document tells you exactly what you need to do to ensure you have robust qPCR data... only then would I look at correlating the data with other data.
Hope that helps.
The real time expression on its own probably is hard in most cases to give an definitive answer to the the solution, such as drug treatment etc. Real time gene expression data, is often backed up either by protein data or phenotypic observations.
As most of the real time expression data performed would be from relative expression (absolute expression is a bit challenge in most cases), means to compare the target gene with one or more house keeping gene expression. While in the case of glucose levels, it is an absolute measurement, so, you can not really just yet draw any conclusion from this method.
As for relative expression, you would need one or more house keeping genes to reference the expression of the target gene such as gene A, then, you would not use ct as your indicator, instead, you can use delta CT or Delta Delta CT or 2^-delta delta CT
It is possible to draw conclusion after you have stable house keeping gene expression value obtained and delta CT value or delta delta ct value calculated providing that your house keeping gene expression is not too far apart from your gene A.
Good luck.
Hi Azeem,
Gene expression values cannot be exactly correlated with the biochemical values and are not always correlated. This is because, mRNA expression values protein expression cannot be same in all the cases.
In post-transcriptional regulation of events, after mRNA is processed, the expression of protein values depends on whether mRNA is translated or not. Since mRNA gets disintegrated after proteins are made, some of the gene regulatory events often withholds mRNA from translation and hence, mRNA is stabilized but prevented from translation. Hence there are situation that higher mRNA levels cannot be exactly correlate with protein expression.
If you are looking for biochemical parameters that are not exactly matching the gene expression, for example, insulin gene expression and glucose levels, there may be different type of correlation may exist.
Hi Azeem,
I agree with the others above and also suggest you look at "The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments" for assistance on qPCR. The document tells you exactly what you need to do to ensure you have robust qPCR data... only then would I look at correlating the data with other data.
Hope that helps.
Hello Azeem - you might wish to look into our recent statistical package (if you do R) for doing just that, called MCMC.qpcr. It takes raw Ct values and figures out correlations with any parameters (or experimental conditions, or their combinations) you might have measured for the same samples. It spits out (and even plots nicely) the effect of each parameter combination on absolute transcript abundance of each gene, with 95%-credible intervals. The tutorial for the package is here: http://www.bio.utexas.edu/research/matz_lab/matzlab/Methods_files/mcmc.qpcr.tutorial.pdf
- Badges
- Science topic