Input Form¶

Use the input form to set up SaddleSum queries. It consists of four sections: term database and weights, statistical parameters, weight processing parameters and output parameters.

The only mandatory parameters are located in the Term Database and Weights section. You must choose an appropriate database from the dropdown box and enter gene labels with their weights. Since all databases that can be selected are in ETD format (see Obtaining Term Datasets), you may enter gene labels as NCBI Gene IDs, as canonical gene names or as aliases. There are two ways to enter weights: through the text box or by uploading a text file. The format is the same in both cases: it consists of several lines where each line should contain a gene label string, followed by spaces or a TAB, followed by a floating point weight value.

The Statistical Parameters determine which database terms are retrieved as significant. These are E-value cutoff, minimum term size, effective database size and statistical method. See Enrichment Statistics of Sum-of-weights Scores for an explanation of SaddleSum statistics and the saddlesum man page for the detailed description of these options. Note that apart from the default Lugannani-Rice statistics, it is also possible to select One-sided Fisher's Exact test statistics, which are based on the hypergeometric distribution. In that case, you must select a cutoff under the weight processing parameters.

The Weight Processing Parameters can be used to easily perform the most common transformations of weights and set weight cutoffs. Again, the saddlesum man page contains all the details about the allowed values.

The look of the SaddleSum output is controlled through the Output Parameters section. You can select the output format (HTML, text or tab), the format of the term relationships graph image (see below) and its color scheme. If you choose text or tab as output format, SaddleSum will not produce a term relationship graph.

To run SaddleSum press the QUERY button.

Output Page¶

To display query results SaddleSum produces an output page showing an image of the term relationships network, the tables listing significant terms, and a query and database summary.

Significant terms¶

are listed in the tables below the term graph. Each namespace with significant terms has its own table. A namespace table shows for each significant term its database identifier, description, the number of associations with genes, score and E-value.

Clicking on term identifier opens a new window or tab in the browser showing the full term description, while clicking on the last three columns brings up a statistical summary of the term and a full list of all genes associated with it together with their weights.

Query and database summary¶

is placed at the bottom of the results page. It echoes most of the input parameters, such as the name of the selected database, statistical parameters and weight transformation. In addition, it shows the statistics collected during the execution of the query that depend on the actual submitted gene identifiers and weights. These include the numbers of submitted and valid gene identifiers, total and used terms, and unused gene identifiers. Finally, you can toggle a list of unrecognized identifiers, if there were any.

Examples¶

The page https://www.ncbi.nlm.nih.gov/CBBresearch/Yu/mn/enrich/examples.html contains links to three examples of collections of weights that could be used with SaddleSum. The links lead to SaddleSum query forms filled with the log2 ratios of gene expressions, taken from selected microarrays from the NCBI GEO database. These are the same weights available as examples for the command-line version (Examples). Once you reach the query forms for an example, you only need to press the QUERY button to run it and obtain the results.