For droplet-based assays, such as digital-droplet PCR or similar, it is common to image/process only a selection of the generated droplets. Conceptually, I would think that increasing the number of droplets that are imaged/processed would increase the assay sensitivity: For low-abundance targets, it becomes more likely that it is included in the 'sampled' population for counting. Is this assumption correct?
I would love to explore this idea more mathematically, but struggling to find any relevant publications and recognize what statistics are at play. I'm particularly interested in learning how increasing the sampling efficiency in droplet-based assays impacts the signal-to-noise and the sensitivity.
Hi,
I assume you are talking about sample utilization. This is of importance for both droplet based and plate based dPCR systems (dead volume is even higher in the 'nanoplate' based systems (40-80%)).
Your sensitivity will indeed depend on:
- the input: If you would like to detect a mutant in a 1: 100 000 ratio, you need at least 100 000 WT copies (actually at least 300 000 to detect this copy with 95% confidence).
(Also, the more volume you screen, the higher your chance to detect something (you can screen for example more than one well))
- sample utilization: if half of your well is not read, you might miss your rare event
...
Depending on your input and the number of droplets, you will be able to calculate indeed the maximal sensitivity (or 'power'). This is something that is done for the Bcr-Abl kit: https://www.bio-rad.com/en-be/product/qxdx-bcr-abl-kit-ivd-ce-ivd?ID=OV9JHH15.
Based on the number of WT copies and the number of droplets you can calculate the maximal sensitivity in that particular well/wells.
Some more general papers to start might be:
- The dMIQE Group, and Jim F Huggett. “The Digital MIQE Guidelines Update: Minimum Information for Publication of Quantitative Digital PCR Experiments for 2020.” Clinical Chemistry 66, no. 8 (August 1, 2020): 1012–29. https://doi.org/10.1093/clinchem/hvaa125.
- Majumdar, Nivedita, Thomas Wessel, and Jeffrey Marks. “Digital PCR Modeling for Maximal Sensitivity, Dynamic Range and Measurement Precision.” PLOS ONE 10, no. 3 (March 25, 2015): e0118833. https://doi.org/10.1371/journal.pone.0118833
- Badges
- Science method