: what are positive and negative controls used for in pcr ? and how do they show that there is a contamination in the stuff ?
In: Other
Assuming you’re asking about polymerase chain reaction.
As positive control, you use a sample known to contain the locus (stretch of DNA) of interest. As a negative control, you use no sample at all — just dilution buffer.
The positive control should return positive (low Ct) and the negative control should return negative (very high/incalculable Ct). If there’s a contamination that causes false positives in one of your reagents, then that’d probably show up as the negative control not returning negative.
Imagine predicting a yes or no answer, there are for possible outcomes of your prediction.
You guessed “yes” and the answer was “yes” (true positive)
You guessed “no” and the answer was “no” (true negative)
You guessed “yes” and the answer was “no” (false positive)
You guessed “no” and the answer was “yes” (false negative)
The reality is, any possible prediction will be tainted with all of the above, so what are you testing?
Let’s pretend it’s a cancer diagnosis.
In this case guessing “Yes” and the answer is “no” just means maybe the person took some extra tests or had a ‘scare’. That’s not terrible.
Guessing “no” and and the answer is “yes” means the person died. That’s muy no bueno.
So there is a sliding scale you’d invent as the predictor, you can “bias the test to always favor “yes”es unless it’s really, really, really, really sure the answer is no. This creates a lot more false positives, but has almost no false negatives. You’re trading expense for lives. A fair trade.
But how do you, as a real life predictor, really know where the test is ‘programmed’ outside of a college level statistics word problem? You’d include some rigged ‘controls’, you already know the answer to those because you rigged it. If the test hits ‘no’ on a lot of positive controls you might risking more false negatives than you’d prefer. If the test hits a lot of ‘yes’s on negative controls, you might be taking a different risk.
If the results are all over the place the test is designed wrong or your haven’t fully evaluated your procedures properly.
Latest Answers