How to use a heatmap to review results from a set enrichment analysis?

After performing a gene set enrichment analysis, you can pick the top results from the results table to be displayed on a heatmap. The heatmap allows you to rank those genes that are co-mentioned the most with the selected concepts.

create_heatmap_from_enrichment

Steps

Note:

A user action, such as click or enter, is italicized.

A visible item on the platform, such as a [link] or a [button], is in square brackets.

A link to another article or external source is underlined.

1. Perform a gene set enrichment analysis by following this article.

2. In the results table, click the top results by selecting individual rows, one by one.

Note: Click the main checkbox in the top left corner to select all concepts shown on the first page.

select_all_pathway_analysis

selection_of_pathway_analysis_results3. Scroll to the bottom of the page and click on the [Show on Heatmap] button located below the results table.

show_on_heatmap4. A heatmap is then generated. If needed, click the [Swap axis] button to switch the the column and row arrangement of the heatmap to better fit the screen.

Note: The heatmap colors are based on number of references (both abstract co-occurrence and database annotations). In the case where there are large outliers, slide the color scale bar to change the range of the reference count values, thus changing the color intensity of the heatmap.

heatmap_reference_scroll_bar

5. Click the [downward arrow] located next to the [Abstract co-occurrence] tab to switch the heatmap from showing abstract co-occurrence to sentence co-occurrence.

Note: When the heatmap view is in the [Sentence co-occurrence] mode, the color intensity of the heatmap reflects the number of references where each pair of concepts for Set 1 (e.g. pathways) versus Set 2 (e.g. genes) co-occurs in a sentence within the abstracts of publications.

6. Click on the a row or a column name in the heatmmap to sort (this sorts the count values from high to low (0r vice-versa) for that row or column).

Note: Another way to sort is to click the [Sorting actions] dropdown box for the different options to sort the heatmap.

sort_heatmap7. Hover over a cell in the heatmap to get the reference count for a pair of concepts and click it to get an initial list of publication and database references where the concepts co-occur. 

click_cell_on_heatmap8. In the pop up window, click [Open full search page] to get the entire search result.

Note: Clicking [Open full search page] brings you to [Search] with the search box already filled out for the pair of concepts. Refer to this article on how to filter search results.

9. Drag your cursor across those concepts your want to select. Various additional actions can taken after selection of the heatmap:

actions_in_heatmap-1

  • To view the concepts in a relation map, click [Show on relation map]. This will open as a separate window.
  • To save the selected concepts as a set, click [Save as set]. The [Save set] pop-up window appears. Click [Save] after entering the appropriate name, tag, and description for the new set. The set will be saved to the [Sets] folder.
  • To delete the selected concepts, rows, or columns, click the [Deleting actions] dropdown box.

10.  Click [Export to CSV] to save the entire heatmap, which will be saved in the format of a count table displaying the reference numbers.

Possible next steps

  • Click through the heatmap to look at those concept relations with high publication counts and explore the search results
  • Select concepts in the heatmap to build a relation map

Related resources