To anyone interested in the issue, I seem to have come across what might be the reason for the diagonal false positives.

Bio-Rad’s manual says that after the PCR, the plate can be left at 12°C or 4°C overnight. I found that it is actually desireable that it is left overnight before reading, as reading it shortly after the PCR finishes greatly reduces the droplet numbers - see figure 1.

I found that leaving the plate for ~5h at 12°C is also sufficient (see figure 2), but if you run a full plate, it’s hard to fit the plate setup, droplet generation, 2h of PCR and 5h of incubation into one working day.

I have no definite proof for what happens during these 5 hours, but I can speculate. When you take the plate out of the cycler just after the PCR is finished, you can see water condensation on the walls of the wells. This condensation disappears if you incubate the plate at 12°C for (apparently) at least 5 hours. The condensation can’t come from anywhere else other than the droplets themselves, which must mean that the droplets shrink in size. Here’s what the Bio-Rad’s manual says about the droplet reader:

The reader measures fluorescence intensity of each droplet and detects the size and shape as droplets pass the detector; droplets are excluded if they do not meet quality metrics.

In regard to diagonal false positives, what I believe happens is that as a large number of droplets are rejected because of being too small. Some of them, however, just about make it pass the size filter, even though they are much smaller than the average dropplet. Because they are smaller, their apparent background fluorescence is stronger. And importantly - it is stronger to an equal extent in both channels, hence they find themselves on the diagonal of the 2D plot.

I let Bio-Rad know about this. They said they got similar reports about getting low droplet numbers if the plate is read immediately from another research group and they said they'd investigate it themselves in their labs.

To anyone interested in the issue, I seem to have come across what might be the reason for the diagonal false positives.

Bio-Rad’s manual says that after the PCR, the plate can be left at 12°C or 4°C overnight. I found that it is actually desireable that it is left overnight before reading, as reading it shortly after the PCR finishes greatly reduces the droplet numbers - see figure 1.

I found that leaving the plate for ~5h at 12°C is also sufficient (see figure 2), but if you run a full plate, it’s hard to fit the plate setup, droplet generation, 2h of PCR and 5h of incubation into one working day.

I have no definite proof for what happens during these 5 hours, but I can speculate. When you take the plate out of the cycler just after the PCR is finished, you can see water condensation on the walls of the wells. This condensation disappears if you incubate the plate at 12°C for (apparently) at least 5 hours. The condensation can’t come from anywhere else other than the droplets themselves, which must mean that the droplets shrink in size. Here’s what the Bio-Rad’s manual says about the droplet reader:

The reader measures fluorescence intensity of each droplet and detects the size and shape as droplets pass the detector; droplets are excluded if they do not meet quality metrics.

In regard to diagonal false positives, what I believe happens is that as a large number of droplets are rejected because of being too small. Some of them, however, just about make it pass the size filter, even though they are much smaller than the average dropplet. Because they are smaller, their apparent background fluorescence is stronger. And importantly - it is stronger to an equal extent in both channels, hence they find themselves on the diagonal of the 2D plot.

I let Bio-Rad know about this. They said they got similar reports about getting low droplet numbers if the plate is read immediately from another research group and they said they'd investigate it themselves in their labs.

Dear Andrzej,

thank you very much for sharing this extremely interesting and help full knowledge!!!

Do think about publishing it? It might be a small paper, but very important. Especially for liquid biopsy, where you always struggle with low amounts of targets!

If you are interested in a comprehensive study we let us talk about it! I can prepare some samples, analyze them with the same batch of primers (for example we can do mRNA analysis with RT one step protocol) and send everything (samples+ primers) to you, so that you can do the same on your machine. Best

This is interesting Andrzej! I was just reading that oil can become "thinner" when heated. Maybe the droplets are more prone for breaking by the droplet reader itself until cooled?

I know this post was from a while ago, but is it possible for you to share a photo of the diagonal droplets? I'm new to dd PCR myself and working out the kinks for detecting a particular gene.

Hi Sarah Johnson,

unfortunately, I no longer work at the company where I had observed those diagonal false positives.

I also heard that the oil temperature can significantly affect the droplets and in hindsight I do remember that there was a brief time when the company had air conditioning issues in the lab where the droplet readers lived, but I can't remember whether the presence of the diagonal false positives correlated with those air con problems.

Martin Horst Dieter Neumann, we recently submitted a technical paper on ddPCR, which will include these results and conclusions.

Andrzej Rutkowski wski Thank you for this topic. I am currently suffering under this issue. Now that my targets are very low. It makes it nearly impossible to quantify accurately.

Sarah Johnson You can link up and ask a few questions that may be a problem for you