I have designed a ddPCR assay to detect gene edits in my organism. The PCR amplicon length is 100 bp, and two probes are within the amplicon. Probe 1 (labelled with FAM) is specific to mutant and Probe 2 (labelled with Hex) as a reference to give an approximate count. When edit is present I expect fluorescence from both the probes. Recently I ran a wildtype sample the reference looks ok however target which I was not expecting anything shows a weird empty bubble pattern. There are two clusters very close to each other. My positive controls are all good.
Fellow Researcher,
After observing the amplitude difference between the two zones/layers of droplets from the FAM channel in your WT sample (1st Image) I believe that the threshold needs to be adjusted between the two zones, slightly higher that the lowest droplet zone.
The lowest droplet zone represents the zero target droplet zone and the upper zone represents a zone of droplets that have a target but probably not the mutant sequence target, which you designed the primers' set for.
The low amplitude difference between dead droplets and mutant droplets could be caused due to a primer specificity problem or an incorrect ddPCR thermal profile.
You could test your WT sample with a thermal gradient experiment, evaluating different annealing/extension thermal profiles. If that doesn't work and you still detect positive droplets in your WT sample then you should check another WT sample in case of contamination. If that doesn't solve the problem then you might consider choosing another primer set.
I hope I understood your case and helped you,
Kind Regards
@Antonis Thank you very much for your response. I agree with you. Yes, my edit is in such a location that it was impossible to design a regular probe. However, I tried and tested them and got two clusters(cross-reactivity) in the edits since there is a single base difference only between edit and WT. I designed the LNA probe and also used DARK PROBES( wt seq probe with 3' phosphate and 3 degrees warmer than the FAM (mutant) probe ) to mask cross-reactivity with the FAM (edit) probe. If I set the threshold based on my edits (8 to 10k) then there seems no problem. Will incresing the concentration of dark probe help?
Hi,
Could you share the 2D plots too? It might be that your assay is ok and that this effect is just due to the positioning of your HEX-positive droplets (that do not contain the FAM-target).
Out of curiosity: is there a reason you are doing a 'drop on' assay, rather then a 'drop off' assay? Are you only interested in the specific edit only?
Caroline Weydert Hello, here is my 2D plot. Yes, I am interested in the specific edits so I am trying a drop on assay. Just for your information, I am using DARK probes too ( wt seq probe with 3' phosphate and 3 degrees warmer than the FAM (mutant) probe) to prevent the Mutant probe from picking up the WT sequence as there is a single BP difference. Thank you
Hi,
You can see that the HEX only population is a bit higher then the double negative population.
If you then look in 1D, this looks like a double line: the double negatives + the droplets negative for FAM, but positive for HEX.
Caroline Weydert I am a bit confused now. The 1D Plot you are referring to is from channel1 which is FAM (probe for mutant) top figure and the bottom one is from channel 2 (Hex REFERENCE). Technically I should only get negative for FAM and positive for HEX in the wildtype strain. I agree that the green arrow you are pointing in 2d plot is for HEX, but the 1D plot you are referring to is from channel 1 (FAM). HEX has a clean separation see figure below (green). Thanks
Hi,
The 2D plot shows FAM in y-axes and HEX in the x-axes.
Your assay has the following characteristics:
- mutant +: double positives
- WT: HEX positive
The 2D plot from the WT indeed shows a nice separation in HEX.
You can also see that, although you only see HEX, the HEX-cloud is a bit tilted up.
This declares why you see 2 populations in FAM (the "weird pattern"): the lower line you observe contains no target, the upper line contains HEX-target, but still no FAM target. This is due to the upwards tilting of the HEX-cloud in 2D.
I added a composition of the full plot for clarification.
This is not a problem, just do your analysis in 2D (please use the drop off algorithm) or place your threshold in 1D above both negative FAM populations. You could also use a TILT algorithm that wil tilt your HEX cloud and probably straighten your plots.
@Caroline Thank you very much for your response. I got it now. Is TILT part of QuantaSoft?
Hi, It is part of QX Manager, it can be downloaded from the website or obtained through your local support.
Caroline Weydert Thank you very much for your help. The analysis was automatically done in auto-tilt mode. The result looks fine. The analysis does not consider that "weird band". However, when I checked my data csv file it shows fluorescence from both the channels (CH1+CH2) it should have been CH- & CH2+ (Hex+). I can only eliminate it if I adjust the threshold above that band. Thank you.
- Badges
- Science method