Hello,
I want to show cell % of cells with a specific mRNA expression. so I did the staining with my targeted protein once with mRNA C and another with mRNA D.
I used Qupath to get the cell number of cells that expressed both mRNA C+D
The cells that expressed only C the cells that expressed only d and the cells with no expression, only DAPI.
mRNA C+ D+
C+D-
C-D+
C-D-DAPI
then I got the averaging and the percentage of each and plotted using Qupath.
I 'm only concerned with mRNA C so I plotted C+ D+, C+D-, and I added (C-D-DAPI + C-D+) and plotted them as the rest of the population.
now my question is that I got an error bar on the rest of the population (C-D-DAPI + C-D+)that my lab mate said that this must not happen.
I don't understand why and what to do to fix it.
Why/how are you labelling proteins with mRNA?
Why are you also looking for a different target (D) that you are not even interested in?
How are you "using Qupath to get the cell numbers"? Are these adherent cells you've taken pictures of, or what? How are you running your segmentation, if so?
And aside from all of these issues, why are you taking such a ridiculously long winded and circuitous route to analysis?
You have a label for your cells of interest (C), and you have a label for all cells (DAPI), you want to know % C+.
You can ignore D+ cells completely: you're not interested in those (apparently).
You run 3-5 biological replicates (at least), and count C+ cells and total cells to get your 3-5 "% C+" values (100x C+/total). Then you present the mean + SEM of these (and ideally show all 3-5 datapoints, too).
You absolutely can have errors for your % C+, and you really should.
Hello,
1. Writing a "protein" was a mistake, I'm using the RNAscop technique to stain specific mRNA.
2. Because I'm interested in knowing the specific cell type my targeted gene is expressed more. I want to know whether is expressed more in progenitor stem cells or immature neurons. so I'm using mRNA probs for mRNA that are known to be specifically expressed in these different cell types.
3. The RNAscop was performed on tissues and then imaged using confocal. The image file was moved to Qupath where I created a single measurement classification for each channel and obtained the measurement.
I'm interested in showing how many cells have both positive signals (C+ D+) and cells with only (C+)
because the purpose is to show that my specific mRNA is expressed in a certain cell type (a cell type expressing D mRNA).
So eventually I need to show
1. cells that express both mRNA
2. cells that only express mRNA C+
3. cells that only express mRNA D+
4. % of cells that don't express either.
My lab mate pointed out that I shouldn't have an error bar on % of cells that don't have expression, which I don't understand and he won't explain.
No, he's wrong, since all your measurements will have error, both due to technical variation and biological variation.
Nevertheless, it seems like you're overthinking this.
If 30% of your cells (+/- 5%) express C, then by extension, 70% of your cells (+/- 5%) do not. You don't need to even show "% of cells that don't express either", since it's instantly apparent from the data you're already showing. Assume your readers can add up to 100.
So just list % of double positives, and % of each single positive (all with respective errors), and...you're done.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Molecular Methods
- Cloning