Submit your RealTime PCR questions and watch the rest of our videos at http://ow.ly/bQh0l. Life Technologies Sr. Field Application Specialist Doug Rains helps with the understanding of baselines in RealTime PCR. We're looking at a fairly standard realtime amplification plot. We have some nice curves, each of which has the familiar geometric phase, linear phase, and plateau phase. So far, so good. But what's all this . . . junk in the early cycles? Well, friends, if you said "junk," you were right. That's right, I said it junk, trash, waste, detritus, garbage, otherwise known as noise.
It's the stuff we see before our actual signal from amplification gets high enough to overcome that noise. And, as the rather impolite adjectives I used a second ago would suggest, it's completely useless to us. This noise does have an effect on our curves. Our job is to minimize that effect by effectively subtracting out the noise. We do that by establishing what's known as a baseline a cycletocycle range over which only noise can be seen, prior to the appearance of curves. Once established, the software will effectively subtract out the noise on a wellbywell basis, greatly improving the quality of our data.
Let's switch the Yaxis to linear scale for a moment to illustrate the effect of baseline subtraction. Here's our data prior to baselining. Note how every sample begins from a slightly different spot on the Yaxis, causing our geometric phase data— this curvy part over here when we're in a linear scale— to look horrible. But once we subtract noise, every sample begins from the same point 0. And as a result, the data clean up nicely.
The value we get after normalizing for background noise is something called deltaRn. If you ever look closely at a logscale amplification curve— the one we're used to seeing— you'll notice that deltaRn is what's graphed on the Yaxis.
But before you go, just note that there are two ways to set baselines in Applied Biosystems® realtime PCR software: manually, and automatically. If you do it the manual way, you set the baseline range under Analysis Settings. You either set it for a single assay, in which case all wells for that assay get the same subtraction . . . or you can go under Advanced Settings and set wells individually.
Better yet, just use the default setting of Auto Baselining. With this selected, the software figures out how much noise needs to be subtracted from each well individually, and, as such, generally produces the best results.
So why have a manual feature? Well, Auto does fail on occasion, especially with some SYBR® Assays and nonstandard chemistries. You'll know auto has malfunctioned by the shapes of your curves. If they look more Sshaped than they should, it could be that auto has misapplied the baseline and set the End cycle too low. As a result, not enough noise is being subtracted, and the curves take on a strange shape. To fix the problem, switch over to manual mode for that assay and raise the End cycle until the curves take on a regular shape.