thebench.bio #1 Ecosystem for the Bio-Tech Industry
All PCR & qPCR questions

Question 17 of 20 Advanced Scenario

A sample gives one late Cq near 38 while its replicates are negative. How do you decide whether the target is really there?

Short answer

At a few copies per well, chance alone decides whether a replicate contains a molecule, so mixed positive and negative replicates are expected near the detection limit. Decide using the assay's validated limit of detection, clean negative and no-template controls, the product's identity and a retest with more replicates or more input, not the single Cq.

Why

Molecules are distributed among wells randomly, following a Poisson distribution. If a sample holds on average one copy per well, about 37 percent of wells receive none and stay negative. One positive out of three is therefore what a genuine low-level sample often looks like.

It is also what a single contaminated well looks like. MIQE 2.0 states that a positive in a validated qualitative assay, without contamination or probe degradation, means the target is present whatever the Cq. It also notes that at 90 percent efficiency the detection limit of about 3 copies is reached around cycle 37.

The evidence that separates the two readings is outside the one well. Were all no-template and extraction blanks clean in this run and in the bench's recent history? Does the late product have the right melt temperature or probe signal?

Does the sample sit above or below the validated limit of detection? For example, a stool sample with one of three wells at Cq 38.2 and clean blanks would be retested in eight replicates with more extract. Five of eight positive at similar Cq supports true low-level presence.

Zero of eight points to a sporadic contamination event.

How to work it through

  1. Check the run's no-template, extraction blank and negative controls, and the lab's recent contamination record.
  2. Confirm the late signal is the intended product by melt temperature, probe signal or sequencing.
  3. Compare the observed level with the assay's validated limit of detection.
  4. Retest from the original extract with more replicates, and with a larger input if the matrix allows.
  5. Report the result as detected below the limit of quantification, with the replicate count, rather than as a number.

An example

A wastewater extract gives Cq 38.4 in one of three wells, the other two undetermined. All blanks are clean. The assay's limit of detection, set at 95 percent positive replicates, is 4 copies per reaction.

The extract is retested in 12 wells with double the input volume. Seven of 12 are positive, with Cq values from 37.1 to 38.9. Seven of 12 positive is consistent with just under one copy per well at the doubled input.

The lab reports the target as detected but not quantifiable.

Where the answer stops holding

Retesting helps only if enough extract remains and the target is stable. A low-level contamination source that is persistent, such as an aerosolized amplicon, can produce repeated late positives that look like real signal. Only blanks run alongside, and ideally a separate clean area, rule that out.

Assays with lower efficiency reach the same copy number at later cycles, so a fixed Cq cutoff does not transfer between assays.

A strong answer covers

  • Poisson sampling makes mixed replicates expected near the limit
  • Validity of blanks and contamination history decides interpretation
  • Confirm product identity of the late signal
  • Retest with more replicates or more input
  • Report as detected but not quantifiable

Common mistakes

  • Applying a fixed Cq cutoff such as 35 to call every later signal negative
  • Averaging a positive well with undetermined wells as though undetermined were a number
  • Calling the sample negative because most replicates were negative