Sort trellised charts by P-value
From Array Suite Wiki
Sort trellised Land charts by P-value
In certain Views within Land data, you can apply a trellis to subset the data across multiple charts by a selected metadata column. Once you have applied a trellis, you can change the ordering of the charts from "Sort by trellis covariate" (alphabetically by the covariate), to "sort by P-value". This enables you to quickly find the charts with data that were found to have significant differences between the data in the visible chart.
Step 1: Open a supported plot
Most charts that 1. Can be trellised by a covariate 2. Have "Show summary information" or "Change Regression Line Setting" (so can calculate a P-value) can be re-ordered by P-value. Examples include Gene FPKM, Comparison Correlation, Integration (Scan All Genes), and Survival plots.
Charts that are pre-trellised, such as the Gene FPKM (Project View) or Transcript FPKM (Individual Chart) cannot be changed.
For example, search for ERBB2 in OncoGEO, then select the Gene FPKM chart.
Step 2: trellis on a covariate
Select a useful metadata column to subset the data on. For example, if looking at disease-vs-normal expression across a variety of diseases, try trellising on TissueCategory. It's best to trellis on fewer than 50 charts, or the sorting in step 4 may take a while.
For example, trellis ERBB2 expression on TissueCategory, to see how ERBB2 expression differs across multiple disease and normal samples in each Tissue Category.
Step 3: Display the P-value
Before sorting on P-value, be sure to display the P-value information.
In certain plots (Scatter plots such as Integration| Scan All Genes), under Customize, you will find "Change Regression Line Setting". Select "Line+Summary".
In other plots (Variable Views, Survival plots), click "Show Summary Information".
Step 4: Click "Sort on P-value"
Click "Sort on P-value" to re-order the data.
For example, change "Specify Multiple Profile Columns" to TissueCategory+DiseaseCategory,
then sort on P-value.
Scroll down to explore the differences across diseases within each Tissue Category.
For example, two different Disease Categories from the same tissue may have very different expression: