Data Analysis on the ABI PRISM
®
7700 Sequence Detection System: Setting
Baselines and Thresholds
Overview
In order for accuracy and precision to be optimal, the assay must be properly
evaluated and a few adjustments need to be made.
There are three important parameters to be assessed:
• Baseline
• Threshold
• Ct value
Data Analysis Tutorial
To accurately reflect the quantity of a particular target within a reaction, i.e. the
amount of PCR product, it is critical that the point of measurement be accurately
determined. Real-time analysis on the ABI PRISM
®
7700 Sequence Detection
System involves three principle determinants for more accurate, reproducible
data.
Baseline Value
During PCR, changing reaction conditions and environment can influence
fluorescence. In general, the level of fluorescence in any one well corresponds
to the amount of target present. Fluorescence levels may fluctuate due to
changes in the reaction medium creating a background signal. The background
signal is most evident during the initial cycles of PCR prior to significant
accumulation of the target amplicon. During these early PCR cycles, the
background signal in all wells is used to determine the “baseline fluorescence”
across the entire reaction plate. The goal of data analysis is to determine when
target amplification is sufficiently above the background signal, facilitating more
accurate measurement of fluorescence.
Threshold
The threshold is the numerical value assigned for each run, which reflects a
statistically significant point above the calculated baseline.
Ct Value
The Threshold Cycle (Ct) reflects the cycle number at which the fluorescence
generated within a reaction crosses the threshold. The Ct value assigned to a
particular well thus reflects the point during the reaction at which a sufficient
number of amplicons have accumulated, in that well, to be at a statistically
significant point above the baseline.
1
Baseline Setting
The first step of data analysis is to accurately set the baseline. By default, the
software sets the baseline from cycles 3-15.
An example of this initial analysis is shown here:
Figure 1: Default Baseline
Default
Threshold
Baseline default is
cycles 3 to 15
Threshold default is
10 standard
deviations above
mean fluorescence
generated during
baseline cycles
To establish if the baseline needs to be adjusted, it is important to determine
which reaction in the assay emerges earliest above the baseline. In order to
determine this, double-click on the y-axis and change the plot to Linear View.
Figure 2: Baseline Adjustment
Initial
Amplification
Change plot to linear view for
easiest viewing
Verify that the reactions are
flat at zero
Determine which reaction
emerges earliest above
baseline
If the reaction emerges after
cycle 15, then no
adjustments need to be made
2
If the first reaction emerges after cycle 15, then no adjustments to the baseline
are necessary.
Adjusting the Baseline
A highly abundant target (i.e., 18s rRNA) will often amplify very early in the PCR
cycles. In this type of assay, it is necessary to adjust the stop value of the
baseline setting. As shown in the example below, the amplification plot displays
the characteristics of a highly abundant target, and reflects the need to adjust the
baseline.
Figure 3: Log View of Early Amplification
To determine the appropriate stop value, double-click on the y-axis and change
the view to linear.
Figure 4: Linear View
3
For a quick check, adjust the Baseline stop value to six and hit the Update
button. Next, determine the first cycle at which amplification is detected.
Adjust the stop value one to two cycles before the earliest amplification. In the
example below the best stop value was determined to be at cycle 8.
Figure 5: Good Baseline Setting
Assess earliest CT.
Adjust stop value to
one to two cycles
before amplification.
In this case the stop
value was determined
to be 8.
Double-click on the y-axis to return to log view.
Incorrect Baseline Settings
The following examples depict what an amplification plot will look like if the
baseline stop value is set too high or too low.
4
Baseline Stop Value is Too Low
The settings in the example below were changed to cycles 3-8. Note that the plot
begins to dip below zero around cycle 14. To fix a plot like this, extend the
baseline stop back out to 15 and click on update calculations.
Figure 6: Baseline Setting Too Low
Initial baseline set too low
Baseline Stop Value is Too High
The settings in this plot were changed from 3-28. The fluorescence is not flat at
zero; therefore the baseline stop is set to high. To correct it, lower the baseline
stop point to a value of 15. View the plot in linear view to assist in determining the
best stop point.
Figure 7: Baseline Setting Too High
5
Threshold Value
Once the baseline has been set correctly, the software will automatically set the
threshold at 10 standard deviations above the mean baseline fluorescence. At
this point, the threshold can be adjusted manually by placing the cursor on the
threshold line and dragging the line to the point on the amplification plot that
follows these parameters:
1. Linear phase of exponential amplification.
The threshold should be placed in the region of exponential amplification
across all of the amplification plots. This region is depicted in the log view
of the amplification plots as the portion of the plot, which is linear. The
threshold line should neither be placed in the plateau phase nor in the
initial linear phase of amplification, which may be too low, and into the
background fluorescence.
2. Precision maximization.
The threshold should be set at a point that maximizes the precision of
replicates. It is common to find that the precision amongst the replicates
increases as the amplification progresses further into the exponential
phase of the reaction. Therefore, to assure maximal precision, the
threshold value must be set within the exponential range above any
background noise within the assay.
3. Sensitivity maximization.
The threshold should also be set to maximize the sensitivity of the assay.
It is important that the threshold be placed at the point which best reflects
all orders of magnitude in the assay(s) across the plate.
The example below depicts the ideal location for setting the threshold.
Note: There will usually be a "window" or range of values within which a
threshold setting will fit the aforementioned criteria and yield optimum
results.
6
Figure 8: Good Threshold Setting
Threshold
Value
Plateau
Range
Exponential
Phase
Background
Fluorescence
The Exponential phase of amplification is the most accurate place to take a
measurement of the amplicon. The Exponential phase of amplification appears
as the middle third of the reaction in the logarithmic view. The default numerical
value assigned for the threshold depicts a value equal to a certain multiplier
(default of 10) above the standard deviation of the mean baseline fluorescence.
The default Threshold value may not be the best setting for a particular assay.
There are two ways of determining the appropriate threshold setting: the
standard curve method and the replicate method.
Standard Curve Method
The figures below depict the default setting for the threshold; it is too low and
close to the background noise. The default threshold is also in a place where the
replicates are not as tight. This may result in impaired precision for quantitative
values
7
Figure 9: Default Threshold Setting
Threshold Value
Threshold Bar
Figure 10: Corresponding Standard Curve Values
Discrimination between
unknown replicates is poor
Slope value is poor
8
Figure 11: Adjusted Threshold
Threshold Value
Threshold adjusted higher
in exponential phase and
below plateau
Threshold Bar
Moving the threshold higher in the exponential phase of amplification, as shown
above, gave rise to more precise results.
Figure 12: Adjusted Threshold
Discrimination between replicates
is good Slope value is improved
Slope value is improved
Correlation Coefficient has
improved
By adjusting the threshold, the Slope and Correlation Coefficient values have
improved.
9
Replicate Method:
When a standard curve has not been run, the replicate method can be used to
properly set the threshold. In the figure below replicates of sample sets are being
compared. Once again the default threshold is too low; The Ct’s of the replicates
are scattered a bit and the background is coming above the threshold.
Figure 13: Thresholds for Replicates
Default Threshold is near
background
Replicates are scattered
and not very tight.
The difference in Ct values
are greater than 0.5
Background
When the X-axis is expanded, it is easier to see just how spread apart the
replicates are. The threshold can be easily set when viewed this way.
Figure 14: Expanded X-axis
Replicates are spread
apart at threshold
Expanded X-axis
10
The threshold can be adjusted to improve the precision of the replicates. After
adjustment, the Ct values of the replicates are much tighter.
Figure 15: Adjusted Threshold
Threshold is set
properly
Replicates are tight,
within 0.5 Cts
Data Analysis Summary
Step 1: Determine the correct Baseline Setting
Step 2: Set the proper Threshold
In most cases for a given assay (i.e. same primer and probe set) use the same
threshold and baseline settings. Using the same settings from assay to assay
will improve the precision and make the data more comparable from plate to
plate and instrument to instrument.
Setting appropriate Baselines and Thresholds will help to assess the quality
of an assay
If threshold and baseline settings are not helping with precision and accuracy,
the assay may not have been properly optimized
Note: The threshold and baseline settings will not be saved. It is best to
record the baseline and the numerical threshold and manually set them the
next time the run is opened or another assay is performed.
Data Capture
After a run is complete, the data can be manipulated for presentation purposes.
After the appropriate settings have been determined, the Experimental Report,
Amplification Plot, Standard Curve and other views can be printed out and filed.
11
The data can also be exported to Microsoft®
Excel or another spreadsheet
program. When the Sequence Detection file is opened and all settings have been
adjusted, go to File > Export > Experimental Report or Results. This will create a
text file, which can be opened in programs like Microsoft®
Excel and
manipulated.
There are several tutorials on-line to help with some manipulations of data in an
Excel like program.
Summary:
The SDS software and instrumentation collects fluorescent data as amplicon
accumulates.
The data collected is displayed in an Amplification Plot that can be viewed in
two ways: Linear and Logarithmic.
The Threshold Cycle (Ct) is determined when the level of fluorescence from
accumulating amplicon crosses the Threshold.
Ct vs. Well position is a good way of determining outliers in your reaction.
Standard curves can be used in the reaction to determine unknown
quantities.
The software displays the standard curve as well as the unknowns in the
Standard Curve view.
The quantities of the unknowns are determined automatically and displayed
in the Experimental Report.
Evaluating the Raw Data and Multicomponent View are ways to monitor the
fluorescence in the reaction as the reaction progresses.
Setting Baselines and Thresholds properly will increase the precision of a
given assay.
© 2002. Applied Biosystems. All rights reserved.
For Research Use Only. Not for use in diagnostic procedures.
Information in this document is subject to change without notice. Applied Biosystems assumes no responsibility for any
errors that may appear in this document. This document is believed to be complete and accurate at the time of
publication. In no event shall Applied Biosystems be liable for incidental, special, multiple, or consequential damages in
connection with or arising from the use of this document.
Applied Biosystems and ABI PRISM are registered trademarks of Applera Corporation or its subsidiaries in the US and
certain other countries.
Macintosh is a registered trademark of Apple Computer, Inc.
All other trademarks are the sole property of their respective owners.
117GU05-01
Part Number 4370923 Revision A
12