You can compare multiple runs to see the ways in which they differ. This step of the tutorial shows you the process using our example data.
Choose Runs and Comparison Paradigm
- Click the Proteomics Tutorial link to return to the main folder page.
- In the MS2 Runs web part, select all three runs.
- Click Compare in the grid header, and then choose how you want to compare the runs.
- ProteinProphet: Compare by protein as chosen by ProteinProphet.
- Peptide: Compare by peptide.
- Search Engine Protein: Compare by protein as chosen by the search engine.
- SpectraCount: Compare based on spectra counting.
- For this tutorial, select the ProteinProphet comparison option.
- Review the set of ways in which you can control the comparison:
- Protein groups to use
- Peptide requirements for those groups
- Compare by run or by fraction
- For each run or fraction, choose when to show the protein
- Accept the default options and click Compare.
View the Comparison Overview Venn Diagram
At the top of the page, you can expand the Comparison Overview
section by clicking the
to see a Venn diagram of how the runs overlap:
Filter for High Confidence Matches
You can filter the list to show only high confidence matches.
- In the Comparison Details section, select (Grid Views) > Customize Grid.
- Click the Filter tab.
- Open the Protein Group node (click the ) and check the box for Prob.
- Then, in the filter panel that appears on the right side, select Is Greater Than Or Equal To and enter 0.8 in the text field.
- Click Save.
- In the Save Custom Grid dialog, ensure that Default grid view for this page is selected, and click Save.
When comparing based on ProteinProphet results, all proteins in all the protein groups are shown. You can sort by group number to determine if a single group has indistinguishable proteins.
Understand the New Comparison Results
Notice that there are fewer proteins in the list now. Based on the filter, the grid will only show proteins where at least one run being compared has a probability meeting your threshold.