Loading Data and Attributes

Data and genomic annotations can be loaded from local files, HTTP URLs, an IGV data server, or a Distributed Annotation Server (DAS).

Load from File

Load data files by browsing for files on the local file system. See File Formats for information about the file formats IGV accepts.

To load data from the file system:

  1. Select File>Load from File. IGV displays the Select Files window.
  2. Select one or more data files or sample information files, then click OK

IGV will display a warning if the file is an un-indexed ASCII-format file over 50 MB.  It is recommended that such files should be indexed or converted to the binary TDF format prior to loading (see section on igvtools).

Load from URL

To load data from an HTTP URL:

  1. Select File>Load from URL.
  2. Enter the HTTP or FTP URL for a data file or sample information file, then click OK.

Notes:

  • For .bam, .tdf, and indexed file formats the server must support byte-range requests.

 

Load from Server

To load data from the IGV data server:

  1. Select File>Load from Server. The Available Datasets window appears:
  2. Expand the tree to see the datasets.
  3. Select one or more datasets. Click the check box to the left of a dataset to select it.
    Warning: Selecting a folder selects all of its subfolders and all of the datasets in those folders. This can potentially be a huge amount of data. To be sure you are loading only the data you want, open collapsed folders and select only the datasets of interest.
  4. Click OK. IGV displays the genomic data.

Load from a Distributed Annotation System (DAS)

IGV provides limited DAS support for "basic features" requests .  To load from a DAS server, enter the DAS source feature URL, with data source name.  Some examples that are known to work for the hg18 assembly are:

  • http://das.sanger.ac.uk/das/cosmic_mutations/features
  • http://das.ensembl.org/das/ens_36_omim_genes/features
  • http://www.ebi.ac.uk/das-srv/genomicdas/das/hydragenevar_eQTL_1mb/features

Note: The segment tag, if present, is ignored.  IGV generates the segment tag as needed.

UCSC DAS Sources

The UCSC DAS server is organized such that all features for a given genome assembly (e.g. hg18) are served from a single feature URL.  Specific tracks are specified by the type tag.  To prevent queries from overwhelming available memory IGV requires that a type tag parameter be included with UCSC DAS urls,  for example the following URL  loads the "dgv" track

  • http://genome.ucsc.edu/cgi-bin/das/hg18/features?type=dgv

To see all available types for a specific assembly the following URL can be used (substitute an assembly id, e.g. hg18, for <dbID>).

http://genome.ucsc.edu/cgi-bin/das/<dbID>/types?

Feature Visibility Window

The DAS protocol does not provide a reliable way to know how many features will be returned for a given request.  To prevent IGV from freezing when loading high-density DAS tracks a "feature visibility window" is used to prevent loading when zoomed out.  The default value for the window is 250 kb.  This can be changed by right-clicking over the track.

Removing Tracks and Attributes

To remove all tracks and attributes:

  • Click File>New Session. This is essentially the same as restarting IGV.

To remove specific tracks, do one of the following:

  • Right-click a track name and select Remove Tracks in the pop-up menu.
  • Right-click an attribute value, which selects all tracks tagged with that attribute value, and select Remove Tracks in the pop-up menu.
  • Control-click track names (Mac: Command-click), then right-click one of the selections and select Remove Tracks in the pop-up menu.

Note:

  • Sequence data is associated with the gene track; therefore, removing the gene track removes the sequence data.
  • You cannot remove individual attributes during a session, but you can hide them.

 Creating a Chromosome Name Alias File

One of the common causes for a data loading failure is a mismatch in chromosome names between the data file and the IGV genome it is being viewed against.  Many Bowtie users report this problem after aligning to the supplied NCBI index files because chromosomes are named by accession numbers in the form: gi|224589811|ref|NC_000002.11|. 

The workaround is to create an alias file in 2-column tab-delimited format.  The first column contains the chromosome name in your data file, for example wig or bam file.  The second column contains the corresponding name in the genome assembly you are viewing (e.g., chr1 for our "hg19" genome).  For instance, the alias file might look like this:


NC_000002.11  <tab>  gi|224589811|ref|NC_000002.11|
NC_000002.12  <tab>  gi|224589811|ref|NC_000002.12|

 

Name the file after the genome with an underscore, the word "alias", and the extension .tab.  For example, hg19_alias.tab.  Place this file in the igv directory. The default location for this folder <user home>/igv/genomes, it can be changed in Preferences -> Advanced

Note: Certain well-known aliases are built into IGV and do not require an alias file.  These include mappings that involve adding or removing the prefix "chr" to the name, for example  1 -> chr1 and chr1 -> 1.    Also, NCBI identifiers that start with "gi|" and follow the pattern illustrated in the example above are automatically mapped.