Hello Researchgate community,
I recently ran into a bizarre (at least I've never thought it could happen). We selected a few DEGs from the scRNA seq dataset and run q-PCR to validate the results (heart tissue).
One particular DEG of interest was DOWNregulated in cardiomyocytes only, but no change in other cardiac cell types, and no change as a whole after combining data from all cells. However the q-PCR data suggested there is a big UPregulation in the whole heart under the exact same experimental condition.
What could be the causes of near-opposite data from RNAseq vs. q-PCR? How should I interpret it, and what's a logical next step here?
Thank you very much!! Any input is much appreciated!!!
Disparities between RNA-seq and qPCR data can be attributed to technical variability, differences in sensitivity and dynamic range, platform-specific biases, differential isoform expression, and biological factors. To interpret these disparities, evaluate the quality and reproducibility of the data, consider gene-specific characteristics, compare with existing literature or datasets, and integrate multiple lines of evidence.
I noticed that you seem to mention tissue-specific expression but seem to compare cell types within tissues which could be a source of variation: "One particular DEG of interest was DOWNregulated in cardiomyocytes only, but no change in other cardiac cell types," then followed by "However the q-PCR data suggested there is a big UPregulation in the whole heart...". Can you confirm that the down-regulation was observed in cardiomyocytes whereas up-regulation was in the whole heart? I've worked with heart samples too, there can be gene expression differences with whole heart vs LV. As you know, there are other cell types (fibroblasts) in the heart as well. Have a look at figure 1-i here (https://www.nature.com/articles/nature11919/figures/1), we can see cardiomyocytes expression miR-34c would contribute the most if they were to use LV/whole hearts. It might also matter where the cardiomyocyte was from, there could be differences in atrial vs LV etc.
For the qPCR data, did you do a reference gene normalization study? There reference gene needs to be stable in your experimental conditions.
Can Justin Kiessling
Hi Can, thank you so much for taking the time and effort. I really appreciate your answer.
Sorry for the confusion, the experiments we performed involved only LV cells/tissue. We observed a DOWN-regulation of gene of interest in LV cardiomyocytes only, not the other cell types, via single cell RNA-seq. Then when we perform qPCR with LV tissue, we saw a big UP-regulation. We should have used isolated cardiomyocytes; unfortunately we didn't have the right equipment to do the isolation at the time. It will be done in the future.
My thoughts before the qPCR experiment was that I would either saw a down-regulation or no changes between groups. The UP-regulation data just mathematically don't add-up to me.
We use 18s as reference gene for normalization, which served our lab well in the past using other disease models. I will look more into whether or not it is stably expressed in the current model I am working on. Great suggestion, THANNK YOU!
Interesting indeed, are you confident your qPCR primer is specific to your gene target? It is also possible that when you use LV tissue for the qPCR analysis, you're getting a signal from other cell types but perhaps this signal isn't present in the single cell analysis as you're sampling much less cell counts? A speculative idea.
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biochemical Techniques
- Analytical Chemistry
- pH