User manual
Droplet Digital
™
PCR Applications Guide | 79
Droplet Digital
™
PCR Tips, Assay Considerations, and Troubleshooting
Fig. 9.1. Cross-reacting probes in a rare mutation detection assay. 2-D amplitude (A),
1-D amplitude (B and C), and histogram (D and E) plots.
such confusion, always classify the droplets of cross-reacting assays while viewing the
2-D amplitude plots. For users operating QuantaSoft
™
software version 1.2.10 or earlier,
appropriate droplet classification using a linear threshold may not be possible without
misclassifying some droplets. Upgrading to QuantaSoft software version 1.3.2 or higher
allows for proper classification using the clustering tools.
Frequency
50
100
150
200
250
0
0 2,000 4,000 6,000 8,000 10,000
12,000
Amplitude
Channel 2 amplitude
Event number
2,000
2,000
4,000
6,000
8,000
10,000
12,000
4,000 6,000 8,000 10,000
0
12,000
0
Channel 1 amplitude
0
Channel 2 amplitude
2,000
2,000
4,000
6,000
8,000
10,000
12,000
4,000 6,000 8,000 10,000
12,000
0
A
Channel 1 amplitude
2,000
4,000
6,000
8,000
10,000
12,000
0
B
0 2,000 4,000 6,000 8,000 10,000
12,000
Event number
C
Frequency
Amplitude
2,000
50
100
150
200
250
300
4,000 6,000 8,000 10,000
0
12,000
0
D E
Probe Cross-Reactivity Can Identify Off-Target Amplification
An unexpected extra cluster of positive droplets with fluorescence intensity less than the
cluster containing the target of interest can be caused by a sequence variant in the target
of interest (Figure 9.2). The droplets that cluster around 10,000 relative fluorescence units
(RFU) contain a variant DNA sequence that is not perfectly matched to the designed probe.
The perfectly matched sequence is the higher cluster around 12,000 RFU.
Often the distinction among two or more positive clusters is desirable because it provides
additional information regarding the sample. If the mid-level cluster represents the detection
of a potentially functional homolog, consider setting the threshold below this cluster to
include it in the quantification, or lowering the annealing temperature so that these two
clusters merge into one cluster. If the mid-level cluster is not desired, consider setting the
threshold above this cluster to exclude it from target quantification.