DSEA

Publications

Napolitano, F., Sirci, F., Carrella, D. & di Bernardo, D. Drug-set enrichment analysis: a novel tool to investigate drug mode of action. Bioinformatics (2015). doi:10.1093/bioinformatics/btv536.

Motivation

A common problem when screening small molecules against a desired phenotypic change is that the screening hits can form an obscure, chemically heterogeneous set of molecules.

As some examples of factors contributing to hide the relevant molecular mechanisms in action, consider the following:

• Structurally similar molecules can bind different targets.
• Structurally different molecules can bind the same target.
• Small molecules that bind the same target can induce different effects.
• Small molecules that bind different targets can induce similar effects.

Method

The Drug Set Enrichment Analysis (DSEA) works on the same principles as GSEA, but using an inverse preparation and interpretation of the data. A set of drugs of interest is tested against a database of pathways. Each pathway in the database is stored as a ranked list of drugs (Perturbagen Expression Profile), sorted from the one most up-regulating the pathway (or gene) to the one most down-regulating it. Current data is based on the Cmap 2.0.

DSEA tend to highlight pathways that are most dysregulated by a set of drugs of interest compared with the other drugs in the database. The importance of a drug-pathway pair is not based on the absolute dysregulation level of the pathway induced by the drug, but rather on the ability of the drug to dysregulate the pathway more than other drugs. This could be the key behind the efficacy of the screening hits and the inefficacy of the other drugs.

Currently supported gene set databases

Pathways analyzed by the DSEA are defined as gene sets collected by various sources, as summarized in the following table:

Running a DSEA analysis

Interpreting results

Results for each database are shown as separate tables, in which pathways are sorted according to relevance, together with the corresponding Enrichment Scores (ESs) and nominal p-values. Top 10% pathways are shown for each database.

P-values assess how much the ranks of the chosen drugs are consistently small (up-regulation) or large (down-regulation) for each pathway. Positive ESs correspond to up-regulated pathways, negative ESs correspond to down-regulated pathways. Note that, due to cell adaptation effects, the direction of dysregulation could be opposite to that directly induced by the drug.

A figure on the right of each table shows a hierarchical clustering of the drugs according to their similarities based on pathway rankings for each database. Similar drugs are those that up-regulate and down-regulate the same pathways.

As soon as the Excel report of the results is ready, the "Export" button on top of the page will become active. Excel files contain all the database results as different sheets, together with additional information including individual drug ranks. While DSEA ranks (obtained through p-value sorting) range from 1 to the number of gene sets in the database, individual drug ranks range from 1 to the number of drugs. If a drug-pathway pair (D, P) has rank 1 then the drug D is the one most upregulating the pathway P as compared to the other 1308 drugs in the database. Importantly, this does not imply that P is the pathway most upregulated by D.

License Terms

Use of DSEA tool is free to academic, government and non-profit users for non-commercial use. Use of DSEA application for commercial use by non-profit and for-profit entities requires a license agreement. Parties interested in commercial use may first obtain a demo version of a stand-alone version of DSEA application, subject to this License, by contacting Diego di Bernardo (dibernardo_at_tigem.it). Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

1. Redistributions of source code must retain the above copyright notice. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
2. The end-user documentation included with the redistribution, if any, must include the following acknowledgment: "This product includes software developed by the Systems, Synthetic and Computational Biology Laboratory of the TeleThon Institute of Genetics and Medicine (TIGEM)". If no such end-user documentation is to be included, this acknowledgment shall appear in the software itself, wherever such third-party acknowledgments normally appear
3. This license does not authorize the incorporation of this software into any proprietary programs.
4. THIS SOFTWARE IS PROVIDED "AS IS," AND ANY EXPRESSED OR IMPLIED WARRANTIES, (INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE) ARE DISCLAIMED. IN NO EVENT SHALL OUR INSTITUTE OR THEIR AFFILIATES BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.