Instruction for
using the Excel macro for quadratic regression analysis:
- Go to “Tools” → “Macro” →
“Security” → “Security Level”, change the security level of your Excel to
“medium”.
- Click on "Excel
macro.xls". Save Excel macro.xls to your computer. If you
have Excel 2000, you need to use Excel 2000 macro.xls.
- Replace the data in “Excel
macro.xls” by copying and pasting your data into the working sheet of “Excel
macro.xls”. The columns of your data should be in the following order: probeset,
the first replication of the first time point, ..., the last replication of
the first time point, the first replication of the second time point,..., the
last replication of the second time point,..., the first replication of the
last time point,..., the last replication of the last time point, description.
Note that the first row should be the names of the covariates, the time point
should be put in the numerical form in order to be used by the macro.
- Click the “RUN” button.
- Put the correct data range in
the box under “Data Range”.
- Change P-Value cutoffs, if
desired. Alpha 0 is the p-value cutoff for the overall model p-value, which is
used to determine the presence of differential expression in the experiment.
Alpha 1 is the p-value cutoff for the p-values of linear and quadratic
effects, which is used to classify genes into different regression patterns.
- Change Worksheet names, if
desired. The default worksheet names are the same as the abbreviations for the
9 regression patterns. For example, sheet “LU” contains genes which are
identified to have linear up-regulated regression pattern.
- Click the “OK” button. The
screen will stop flashing when the analysis is finished. Nine additional
sheets will appear, each containing genes that are classified to have one
regression pattern.