GEO Data Sets
=============

![GEO Data Sets widget icon](icons/geo-data-sets.png)

Provides access to data sets from gene expression omnibus ([GEO
DataSets](http://www.ncbi.nlm.nih.gov/gds)).

Signals
-------

**Inputs**:

-   (None)

**Outputs**:

- **Data**

 Data set selected in the widget with genes or samples in rows.

Description
-----------

**[GEO DataSets](http://www.ncbi.nlm.nih.gov/gds)** is a data base of gene
expression curated profiles maintained by [NCBI](http://www.ncbi.nlm.nih.gov/) and included in the [Gene
Expression Omnibus](http://www.ncbi.nlm.nih.gov/geo/info/datasets.html). This Orange widget provides
access to all its data sets and outputs a data set selected for further
processing. For convenience, each dowloaded data set is stored locally.

![GEO Data Sets widget](images/GEOdataset-stamped.png)

1. Information on the GEO data set collection. Cached data sets are the ones currently stored on the computer.
2. Output features. If *Genes or spots* is selected, genes (or spots) will be used as attributes. Alternatively samples
   will be used as attributes. *Merge spots of same gene* averages measures of the same gene. Finally, in the
   *Data set name* you can rename the output data. GEO title will be used as a default name.
3. If *Auto commit is on*, then the selected data set will be automatically communicated to other widgets. Alternatively,
   click *Commit*.
4. *Filter* allows you to search for the data set. Below you see a list of GEO data sets with an ID number (link to the NCBI
   Data Set Browser), title,
   organism used in the experiment, number of samples, features, genes, subsets and a reference number for the PubMed
   journal (link to the article abstract).
5. Short description of the experiment from which the data set is sourced.
6. Select which *Sample Annotations* will be used in the output.

Example
-------

**GEO Data Sets** is similar to the **File** widget. In the example below
we selected *Caffeine effect: time dose and response* data set from the GEO data base and used *Genes or spots* as
attributes. We inspected the data in *Data Table*. Then we selected
3 genes in the **Select Columns** widget for a detailed analysis in another data table.

![](images/GEODataSets-Example2.png)