The
Partial Least Squares Normalization
process
normalizes
data by fitting a
partial least squares
(PLS)
model
to
class variables
and subtracting the predicted fit. This method is appropriate when the class variables represent unwanted effects in the data, for example a batch or day effect, and you want to remove this effect from the data before proceeding with other analyses.
Caution
: If the class variables are confounded with other effects, the differences in the data due to these other effects can also be removed.
The first data set, the
Input Data Set
, contains all of the numeric data to be analyzed. This data set must be in the
tall
format where each sample corresponds to one row and each column corresponds to a separate experimental condition or array.
The
drosophilaaging.sas7bdat
data set, shown below, is a normalized data set derived from the
Drosophila
Aging experiment described in
Sample Case Studies
. It has 49 columns and 100 rows corresponding to 49 arrays and 100 individual
probes
, respectively.
The second data set is the
Experimental Design Data Set (EDDS)
. This required data set tells how the experiment was performed, providing information about the columns in the input data set. Note that one column in the EDDS must be named
ColumnName
and the values contained in this column must exactly match the column names in the input data set.
The
drosophilaaging_exp.sas7bdat
EDDS is shown below. Note that the
ColumnName
column lists the column names in the input data set. The
Array
column corresponds to an
index variable
. Note the
variables
describing experimental conditions.
The
drosophilaaging.sas7bdat
and
drosophilaaging_exp.sas7bdat
data sets are included in the
Sample Data
folder.
For detailed information about the files and data sets used or created by JMP Life Sciences software, see
Files and Data Sets
.
Refer to the
Partial Least Squares Normalization
output documentation for detailed descriptions of the output of this process.