When a sample or a group of sample serves as control baseline in the experiment design, other samples can be normalized by subtracting or divided by the baseline sample(s), e.g. in PCR data to obtain delta delta CT values.

The Normalize to baseline option is available at the Normalization and Scaling section in the context-sensitive menu (Figure 1) upon selection of any data node containing matrix of measurements of observations (e.g. samples) on features (e.g genes).

Figure 1. When a data node containing matrix is selected, Normalize to baseline becomes available on the context sensitive menu

There are three options to choose baseline samples:

• use all samples
• use group
• use matched pairs

Use all samples to create baseline

To normalize data to all the samples, choose to use mean or median of all samples for each feature, and select subtract baseline or ratio to baseline in the normalization method (Figure 2), click Finish.

Figure 2. Use all sample to create baseline to normalize the data

Use a group of sample to create baseline

When there are a subset of samples serve as baseline in the experiment, choose use group as baseline samples.  The specific group needs be specified from the sample attributes (Figure 3)

Figure 3. Use a subgroup of samples to create baseline to normalize the data

Choose this option, select the attribute containing the baseline group information, e.g. Treatment in this example, the samples labeled as Control will be used as baseline. After normalization, the control sample can be filtered out in the report by selecting Remove baseline samples after normalization check button

Use matched pairs

In a paired experiment design, when normalized one sample to its own control – typically there is only one sample serve as a control to its pair (Figure 4), an attribute on the pair information should be specified in addition to baseline group attribute.

Figure 4. Normalize paired sample to its control

Choose the attribute containing information to pair the samples, and select the attribute with control sample information. After normalization, the control sample should be either 0 or 1 depending all the normalization method chosen, so we highly recommend remove control samples in the report.

Your Rating:

Thanks for voting!

Please Wait

Results:

1

rates