PSAT User's Manual

5. Utilizing Tool Options

Modifying the input options can affect the results that are returned by the tool. Knowing how to utilize the different options can help you take full advantage of the tool for performing a useful comparative analysis.

Modifying BLAST Alignment Score Thresholds

1. Scroll to the top of the Genomic Neighborhoods page and select the Edit Settings link in the header of the homoologs table.
2. You should now be back to the input page, and the fields should be pre-populated with the current settings.
3. Select the first BLAST alignment score thresholds option (e-value < 1e-172, bit score > 600, % identity > 90) and resubmit the form.
4. The Genomic Neighborhoods page should reappear with different results. Since you selected to use stricter BLAST scores, the query returned fewer results. As you might expect, homologs were only found in the other Francisella tularensis genomes when using the stricter criteria

Specifying a Subset of Comparison Genomes

1. Select to Edit Settings again.
2. Edit the Comparison Genomes to include the Salmonella, Shewanella, and Vibrio genomes. Find these organisms in the top selection box. When you click on one, an option in the second selection box to select All genomes within the genus is available. Click on this option to select all and note that the third selection box indicates the selection. (select multiple items using Ctrl-Click or Apple-Click).
3. As expected, the Genomic Neighborhoods page shows that no matches were found for these comparison genomes. Select to edit the settings again so that we can try using more relaxed criteria.
4. Select the third, least strict BLAST alignment score thresholds option and resubmit the form.
5. The Genomic Neighborhoods page should reappear again, but now with 16 results. Since you selected to use less strict BLAST scores, the query returned results this time. Note that the results are ordered by homolog clustering score and are grouped by bacterial genus.

Limiting the Genomic Regions Displayed in the Graphic

1. Assume that you are interested in comparing the genomic neighborhoods of homologs in the Shewanella genomes only. Deselect all the homologs in the table by clicking on the toggle checkbox icon . Then reselect each homolog in the Shewanella genome individually.
2. Click the Update Graphic button. Note that only the Shewanella genomes now appear in the graphic.

Finding Genes by Product Description

1. Select to Edit Settings again.
2. Remove 'pilQ' from the Gene name text box and enter cytochrome into the Description text box.
3. Resubmit the form.
4. A list of genes should be displayed with Product including the string 'cytochrome'.
5. Select the gene cydB in the list of results.
6. Note that the total number of hits total is greater than the number of genomes (indicating there may be duplicated regions in some genomes).
7. Click the 'Genome' column header at the top of the results table to sort by genome. Deselect all checkboxes for all hits by clicking the toggle checkbox icon at the top of the results table.
8. Note that the first 2 hits are in the same genome. Select the checkboxes for these 2 hits and scroll down to select the Update Image button.
9. View the genomic neighborhoods graphic for the 2 results

Viewing Top Hits Only

1. Select to Edit Settings again.
2. Under option 4. Select Display Options, change the setting under Limit results based on top hits to select Show only top hits for each comparison genome. Resubmit the form.
3. The results will now only show the top hit for each comparison genome (the hit with the lowest e-value will be considered the 'top' hit).

Filtering Homologs Based on Homolog Cluster Score

1. Select to Edit Settings again.
2. Change the display setting to Only show hits with a homolog clustering score of at least '1' and resubmit the form.
3. Note that the number of hits has increased with the cluster score constraint removed.
4. Browse the table of results, noting the scores for each hit. You will notice that some scores have a value of 1. These results are not filtered based on clustering of genes and so may contain more spurious hits.

You have now completed the tutorial. Continue on to the next section to view some examples of how the tool can be used to help with genome annotation and genomic comparison studies.