PEAKS Filter Properties

As this is a new feature to PEAKS, we want to ensure all interested parties have step by step instructions on how to get the best results from their data.

The filter pane to be discussed below is just one of the many ways PEAKS allows you to controll your results.

Protein View: Note re-ordered and hidden columns

Protein View: Column Order and Visibility is maintained between the Protein View and the report.

Filter Pane: Displaying filters that allow only de novo peptides used for Protein ID (PID): peptides with a score greater than 50 and proteins with more than one supporting peptide.

Results from PID or inChorus may be manipulated and exported in a number of ways. First, we can display de novo peptides, database peptides, and proteins in three separate views. Now is a good time to note that individual columns can be displayed or hidden in each of the three views. Your settings will be preserved between different sessions and results can be sorted by clicking on the column headers in the de novo and peptide views. Each of the de novo, peptide and protein views can be exported to HTML and XLS (Excel) reports and these reports will show only those columns that are visible in each of the views. The details shown for supporting peptides in the Protein Pane for a selected protein in the Protein View will reflect the columns in the Peptide View. For example, the headers selected in the Peptide View will be the same as the headers in the Protein View Peptide Pane.

The filter pane allows you to manipulate and re-organise your results before exporting them. The various parts of the Filter Pane are described below:

In order to understand how the filter pane works, it is easiest to re-load the data file that was created in the Features Walkthrough. Loading up the data, we can see in the Protein View that PEAKS managed to identify two protein clusters: one strong cluster for lactoglobulin beta and one cluster of proteins based on one peptide.

Let's assume that we are only interested in Lactoglobulin Beta and that the other Protein cluster is completely superfluous. So we'll go to the filter pane and select the filters that show only results pertinent to Lactoglobulin Beta. Go to the Possible Filters listing and add the PID filter (which filters out de novo peptides that did not contribute to database peptides) and the Desc filter (Description filter) for proteins. For the description filter, add the regular expression ".*beta.*" (which will filter out all proteins that do not have the word "beta" in them). Then select the second de novo mode, "De novo view shows all peptides that are not filtered". Hit the apply button. Now when we go to the peptide and de novo views, we will see only peptides that contribute to our result.

Protein View: before filters are applied

Selected Filters

Protein View: after filters are applied

Protein View:Only proteins with the word "beta" are shown

What happens if we want to further filter our results? Let's go back to the filter pane and add a peptide filter that shows only database peptides with scores greater than 50. Go to the Possible Filters and select the score filter from the Peptide Filters branch. Remember to select "greater than" and enter 50 in the appropriate filter after adding it to the Active Filters list. The filter can be edited while in either listing and can be added multiple times (in case, for some reason, you only wanted peptides with a score greater than 50 but less than 90). Hit apply filter and let's examine the results.
All peptides with scores less than 50 are removed both from the Peptide View and the Protein Pane within the Protein View.

Protein View: with Protein Pane for gi|229460

After applying filters, unique peptides 2,3,9,17,18,20 are removed due to their scores and the protein stats are updated

There are a couple of settings for the de novo pane. Make sure De novo view shows all peptides that are not filtered is selected and hit apply. Go to the De novo view and note that only 16 out of 22 peptides are shown since 6 peptides (as noted above) have been filtered out due to their low scores. Go back to the Filter Pane, select Remove de novo peptides with no database hits and hit apply again. Go to the De novo view and note that only 12 peptides remain. Four peptides were removed since they did not contribute to any database peptides. Go back to the Filter Pane, select De novo view shows peptides that could not be explained by peptides from the Peptide View and hit apply again. Go to the De novo view and observe that ten peptides are showing. These are the the six de novo peptides that correspond to database peptides that were removed due to our filtering and the four peptides that did not belong to any database peptides in the first place.

All non-filtered de novo peptides

Non-filtered de novo peptides that contributed to the database search

Peptides that could not be explained by lactoglobulin beta

	News \| Products Downloads Support User Comments Corporate
De Novo Sequencing Protein ID PTM Seq. Tag Search Raw Data Tech Support Request Prices Try PEAKS Our Research User Comments	Superior de novo sequencing and protein ID Software PEAKS Manipulating Protein ID Results (Assistance) As this is a new feature to PEAKS, we want to ensure all interested parties have step by step instructions on how to get the best results from their data. The filter pane to be discussed below is just one of the many ways PEAKS allows you to controll your results. To view full screen shots click on the images below: Protein View: Note re-ordered and hidden columns Protein View: Column Order and Visibility is maintained between the Protein View and the report. Filter Pane: Displaying filters that allow only de novo peptides used for Protein ID (PID): peptides with a score greater than 50 and proteins with more than one supporting peptide. Results from PID or inChorus may be manipulated and exported in a number of ways. First, we can display de novo peptides, database peptides, and proteins in three separate views. Now is a good time to note that individual columns can be displayed or hidden in each of the three views. Your settings will be preserved between different sessions and results can be sorted by clicking on the column headers in the de novo and peptide views. Each of the de novo, peptide and protein views can be exported to HTML and XLS (Excel) reports and these reports will show only those columns that are visible in each of the views. The details shown for supporting peptides in the Protein Pane for a selected protein in the Protein View will reflect the columns in the Peptide View. For example, the headers selected in the Peptide View will be the same as the headers in the Protein View Peptide Pane. The filter pane allows you to manipulate and re-organise your results before exporting them. The various parts of the Filter Pane are described below: Possible Filters: This is a listing of the filters, currently available, that can be used to build a particular filter set. Active Filters: This is a a listing of the filters, currently selected, that will be used to filter your results. Add Filter Clone: The currently selected filter in the Possible Filters listing to the Active Filters. Apply Filters: Any/all selected filters will not be used until this button is actually pressed. Edit Filter: Edit the settings for the currently selected filter. Filters in both the Possible Filters and Active Filters can be edited. Options: This lists settings that fundamentally alter how filtering behaves. De novo view shows peptides that could not be explained by peptides from the Peptide View In this mode, the De Novo view only shows peptides that meet one of two conditions. Either they have no corresponding database peptide (and thus presumably describe some new protein not covered in your existing database) or their corresponding database peptide has already been removed by a filter. The de novo peptides still have to pass the filters defined for the de novo view. De novo view shows all peptides that are not filtered In this mode, the De Novo view shows all peptides that have not been filtered, regardless of whether they have corresponding database peptides or not. Remove de novo peptides with no database hits This option removes all peptides that have no corresponding database peptide hits. Selecting the second de novo view mode and this checkbox shows essentially the "reverse" of the first de novo view mode. Save As, Delete Set, Saved Parameters Save and re-use your filters between different sessions of PEAKS. All active filters and de novo view settings are saved. Let's try this out with an example In order to understand how the filter pane works, it is easiest to re-load the data file that was created in the Features Walkthrough. Loading up the data, we can see in the Protein View that PEAKS managed to identify two protein clusters: one strong cluster for lactoglobulin beta and one cluster of proteins based on one peptide. Let's assume that we are only interested in Lactoglobulin Beta and that the other Protein cluster is completely superfluous. So we'll go to the filter pane and select the filters that show only results pertinent to Lactoglobulin Beta. Go to the Possible Filters listing and add the PID filter (which filters out de novo peptides that did not contribute to database peptides) and the Desc filter (Description filter) for proteins. For the description filter, add the regular expression ".beta." (which will filter out all proteins that do not have the word "beta" in them). Then select the second de novo mode, "De novo view shows all peptides that are not filtered". Hit the apply button. Now when we go to the peptide and de novo views, we will see only peptides that contribute to our result. Protein View: before filters are applied Selected Filters Protein View: after filters are applied Protein View:Only proteins with the word "beta" are shown What happens if we want to further filter our results? Let's go back to the filter pane and add a peptide filter that shows only database peptides with scores greater than 50. Go to the Possible Filters and select the score filter from the Peptide Filters branch. Remember to select "greater than" and enter 50 in the appropriate filter after adding it to the Active Filters list. The filter can be edited while in either listing and can be added multiple times (in case, for some reason, you only wanted peptides with a score greater than 50 but less than 90). Hit apply filter and let's examine the results. All peptides with scores less than 50 are removed both from the Peptide View and the Protein Pane within the Protein View. Protein View: with Protein Pane for gi\|229460 After applying filters, unique peptides 2,3,9,17,18,20 are removed due to their scores and the protein stats are updated There are a couple of settings for the de novo pane. Make sure De novo view shows all peptides that are not filtered is selected and hit apply. Go to the De novo view and note that only 16 out of 22 peptides are shown since 6 peptides (as noted above) have been filtered out due to their low scores. Go back to the Filter Pane, select Remove de novo peptides with no database hits and hit apply again. Go to the De novo view and note that only 12 peptides remain. Four peptides were removed since they did not contribute to any database peptides. Go back to the Filter Pane, select De novo view shows peptides that could not be explained by peptides from the Peptide View and hit apply again. Go to the De novo view and observe that ten peptides are showing. These are the the six de novo peptides that correspond to database peptides that were removed due to our filtering and the four peptides that did not belong to any database peptides in the first place. All non-filtered de novo peptides Non-filtered de novo peptides that contributed to the database search Peptides that could not be explained by lactoglobulin beta Summary To sum up, let's note a few facts about how the filters in PEAKS Studio work. Filters are applied sequentially, the result is that a peptide or protein has to pass all filters in order to be displayed. Filter sets can be saved and re-used between sessions, but remember to hit apply filters to use them. While De Novo View shows peptides that could not be explained by peptides from the Peptide View mode, the results of filtering on the De Novo View are "isolated" from the Peptide and Protein View. That is to say, while the removal of a protein will cause its associated de novo peptides to appear in the De Novo View, filtering them out again will have no effect on the Protein View. Use this mode if you are interested in de novo proteins and if you wish to focus on the de novo proteins that could not be explained by proteins in a database. That is, a true de novo peptide/protein. De Novo View shows all peptides that are not filtered mode, the results of filtering cascade fully upward and downwards. For example, if a protein is removed, its associated database and de novo peptides are removed, assuming they are not shared. If a de novo peptide is removed, its associated database peptides and proteins are removed, again with the same assumption. Use this mode if you wish to remove de novo sequences that likely arise from contaminants and unreliable protein results.