ENCODE Resources and Frequently Asked Questions

Response:
Please avoid loading GEO data as a custom track! Rather, as nearly all ENCODE data at GEO are already hosted as tracks on the UCSC browser, load the existing corresponding track.

Take note of the GEO sample accession (GSM) number and enter it into the Track Search tool accessible from the left side of the ENCODE portal page by clicking Search, for example GSM999240. Or use the Advanced Track Search page and select "GEO sample accession" from the pull down menu displaying "Cell, tissue or DNA sample". Click the box next to your track resulting from the search and the "View in Browser" button.

If you have data that is not already in the browser we recommend converting your BED files to bigBed format. You could download our source tools for converting from BED to bigBed (as described in the previous link) or use the tools at the Galaxy website. For questions regarding Galaxy you will have to contact them directly.

ENCODE CELL TYPES

Question:
Which cell types are used by ENCODE?

Response:
On the left side of the ENCODE portal page under Human and Mouse are links to the cell types used in all ENCODE experiments.

ENCODE PROTOCOLS

Question:
Where can I find the ENCODE growth protocol for a specific cell type? For example RCC 7860?

Response:
To find a specific protocol, for example for human RCC 7860 cells, from the ENCODE portal navigate to the Human Cell Types page. Under the "Documents" column for RCC 7860, click the link to connect to see the growth protocol named after the lab that provided the document, in this case "Crawford".

Another path to ENCODE protocols is from the link /ENCODE/protocols/. Navigate to the cell protocols and then human directories to find the link to the same RCC 7860 protocol file as linked on the above Human Cell Types page.

If you have further questions about a protocol contact the lab that registered the protocol.

TRANSCRIPTION FACTORS

Question:
Has transcription factor xxx been mapped by ENCODE?

Response:
A quick way to view the list of transcription factors mapped by ENCODE is to view the ChIP-seq matrix for either human or mouse. Targets are listed horizontally across the top, indicating available mapped transcription factor data. Clicking on the green highlighted boxes will bring you to experiment data specific to the corresponding cell type and target.

Another option is to use the Track Search or File Search tools and to search the "Antibody or target protein" field to see if the desired transcription factor is listed.

MAPPING A CUSTOM TRACK TO TRANSCRIPTION FACTORS

Question:
How do I find overlaps between my own ChIP-seq regions and available ENCODE transcription factors?

Response:
By using the Table Browser tool you can add your ChIP-seq information as a custom track and then use the "intersection" feature to intersect the Txn Factor ChIP track table listed under the Regulation group with your custom track. Note, your custom track should contain ChIP-seq regions in BED format, for more information visit our custom tracks page.

If you are unfamiliar with the Table Browser, please refer to our help page and the section on intersecting data.

FILES NAMED xxxV2

Question:
What is the difference between a file xxx and the related file xxxV2? Why is the xxx file not displayed in the browser?

Response:
For files named similar to xxxV2, often the "V2" refers to a second version that revokes earlier versions that are therefore not displayed in the browser. Revoked files are still available for download, but they will be indicated as "replaced " or "revoked" in the related metadata file named "files.txt" present in the corresponding download directory.

ENCODE METADATA AND FILENAMES

Question:
How do I extract information about an ENCODE experiment from the filename?

Response:
This is not recommended. While ENCODE filenames have some metadata embedded, the information there is not complete nor easily extracted. Rather, use the file's metadata, for example in "files.txt", or access metadata in the following places:

Track Search or File Search

Table Browser tool

public MySQL database

The metadata uses controlled vocabulary (cv.ra), which can be downloaded as a text file here.

ENCODE FILE FORMATS

Question:
How do I learn more about different ENCODE file formats? For example what is the difference between a file.bed and a file.bed9 in the ENCODE methylation data?

Response:
By clicking the File Formats link from the ENCODE portal page you can reach a list of various file formats used in ENCODE. Every ENCODE file has metadata included under a "files.txt" file in the related downloads page. For example, from the HudsonAlpha DNA methylation download page, in the files.txt file, a line after the specific bed9 file in question, wgEncodeHaibMethylRrbsAg04449UwstamgrowprotSitesRep1.bed9, reads 'objstatus=replaced'. This metadata indicates this bed9 file was preliminary data that has since been replaced. A similar note in the automatically displayed README file states: "WARNING - Revoked and replaced data files may be present in this directory."

ENCODE SCORE DEFINITION

Question:
What is the definition of "score" in ENCODE tables?

Response:
The score (between 0-1000) determines how darkly an item is displayed in the browser (with 1000 being black). The darkness of an item's box is proportional to the maximum signal strength observed in any cell line.

To find out exactly how score has been calculated for a specific track, contact the lab that created the data. There are often several links to authors' labs in the credits section for each track at the bottom of a track's description page.

ENCODE DATA IN BED FORMAT

Question:
How do I download ENCODE histone data in BED format? From the Table Browser I can select to download the file in BED format, but I am limited to just a few thousand lines. When I looked in the ENCODE Downloads directory I could only find the path to a bigWig file, for example wgEncodeBroadHistoneGm12878H3k27acStdSig for human build hg19.

Response:
The ENCODE BED files you are looking to download are the 'peak calls', which are in the extended broadPeak or narrowPeak formats, described here. For example, within the database mentioned (H3K27ac histone mark in GM12878 cells) there is a BED representation in the file: "wgEncodeBroadHistoneGm12878H3k27acStdPk.broadPeak.gz". Using the File Search tool you can use the setting "Data Format: Peaks Broad" to narrow your results to only these types of files.

ENCODE BED FILE FORMAT

Question:
What does the name column represent for DNase clustered BED files? I downloaded the ENCODE BED file wgEncodeRegDnaseClustered.bed from the DNase footprinting assay. However, I am having trouble understanding the 4th column in this file. Usually this column, as I understand from the file format FAQ page, is assigned to name.

Response:
For the DNase cluster BED files, the name field represents the number of items in the cluster. To find out more information about each cluster, you can click on the item in the browser image and it will take you to a details page that will list all of the items in the cluster and the cell lines. Here is an example of a details page for a DNase item on chromosome 21. There are 58 items in this cluster and you can see the name value is 58.

ENCODE ChIP-seq BED FILES

Question:
How do I find the meaning of a column of a BED file? I have downloaded ENCODE Chip-Seq BED files that have the following format:

chr21 9825311 9827738 . 1000 . 4.51792 256.60845 261.34671 1809

What is the meaning of the information from the fourth field forward?

Response:
ENCODE has a number of ENCODE-specific formats. ENCODE ChIP-seq files are typically stored in the ENCODE narrowPeak format. This format extends BED6 to include fields for signalValue, two measurements of statistical significance (pValue and qValue), and the offset of a single base 'point source' peak within the region. The dots are used for name and strand which are not applicable.

DOWNLOAD ALL ENCODE DATA

Question:
Is there a service providing ENCODE data on a hard drive? What is the total data volume? We have been trying FTP, but it takes too much bandwidth and time.

Response:
The total volume of ENCODE data are greater than 31 TB. Unfortunately, it is not possible for you to obtain a disk copy, however, there is a new protocol to try called UDR (UDT Enabled Rsync). UDR provides users much faster download rates.

Here is an example using UDR, once installed, to download all the mouse mm9 ENCODE information:

$ udr rsync -avP hgdownload.cse.ucsc.edu::goldenPath/mm9/encodeDCC/ /my/local/mm9/

Please read more about the new UDR method here.

For those not downloading high amounts of data, we highly recommend using rsync. For example:

$ rsync -a -P rsync://hgdownload.cse.ucsc.edu/goldenPath/hg19/encodeDCC/wgEncodeDir/wgEncodeFile ./

Using rsync has the advantage of starting up where it left off after a failure, when ran again.

ENCODE PAPERS

Question:
Where can I find ENCODE papers? I would like a list of the principal ENCODE Papers, can you send a link to a list of a core 30 papers detailing ENCODE's results?

Response:
References to the ENCODE analysis publications of September 2012 can be found here: ENCODE Analysis Package Publications. There is also a comprehensive set of ENCODE-related publications listed on the Publications page linked on the left side of the ENCODE portal page.

CONVERTING WIG FILES TO VARIABLESTEP

Question:
Can I convert WIG files into a variableStep format to use with SitePro? I am trying to use a tool called SitePro within Cistrome. This tool uses WIG and BED files to compute score profiles on the BED regions. I have downloaded, through Cistrome/Galaxy, the ENCODE WIG files which have BED-like structure:

chr1 3002700 3002800 0.17

However, this WIG file's BED-like structure is not accepted by SitePro. Is there a way to format the WIG files as variablestep and not BED-like?

Response:
There is not a way to convert formats using the Genome Browser directly, but you could convert formats using a script. There is an example script in our genomewiki, here.

UNIQUE ENCODE DATA DETAILS

Question:
What does xxx mean in a file in hgdownload/encodeDCC/hg19/wgEncode(track)? For example downloadable files in the wgEncodeCaltechRnaSeq/ directory have a gene_id format like gene_id "GM12878-rep1.1045777" where the first part is the cell type. Would you know what does the last number 1045777 means?

Response:
At the top of the page for each of the download directories you are visiting there is a README.txt file that is automatically displayed. A link is provided that will bring you to a user interface enabling filtering of files by cell type and other parameters, as well as including additional information such as release status, restriction dates, track description, methods, and metadata that can answer such questions.

For example in the README.txt file displayed at the top of the page in the Caltech RNA-seq directory you can find the following link: "http://genome.ucsc.edu/cgi-bin/hgFileUi?db=hg19&g=wgEncodeCaltechRnaSeq"

By navigating to the page above, Caltech RNA-seq Downloadable Files, you can scroll to the bottom (or click the "Description" link in the top right corner) and read the track description's "Methods" section. In the "Data Processing and Analysis" section there is information explaining how the numbers in gene_id, "GM12878-rep1.####" represent de novo identifiers output by Cufflinks software. At the very bottom of the page is a "Credits" section where contacts are listed. You should send remaining process-specific questions about the data you are investigating to the appropriate contact listed.

ENCODE PROTOCOLS

Question:
Which cell protocols were used in my track of interest? Did the Open Chromatin ENCODE tracks use standard ENCODE cell protocols?

Response:
Standard growth protocols were used for all ENCODE experiments, including the Open Chromatin ENCODE tracks. A directory of all ENCODE protocols is available here: http://genome.ucsc.edu/ENCODE/protocols/.

ENCODE GRAPHIC

Question:
May I use the the ENCODE figure from your homepage? I am writing my PhD thesis and I would like to use it in both electronic and printed form.

Response:
The ENCODE graphic displaying how investigators employ a variety of assays and methods to identify functional elements can be used in publications as long as credit is given. Please credit Ian Dunham at EBI and Darryl Leja at NHGRI as noted below the image.

ENCODE TABLE SCHEMA

Question:
How are the columns signalValue and peak calculated in narrowPeak files? For example, I want more information about UW Histone "wgEncodeUwHistone...PkRep1.narrowPeak" files.

Response:
The File Format FAQ provides explanations about various file formats. Also a file's related Track Description page may include important information, such as the UW Histone "Methods section", which describes how data were processed to produce peaks. In the "Credits" section there is also a lab contact. To request further information for UW Histone data, for example, you could contact the lab to learn more about the peak calling algorithm or other methods involved.

When using the Table Browser there is a "describe table schema" button that gives information similar to that located in the File Format FAQ, plus the related Track Description.

For example with settings "group: Regulation", "track: UW Histone", and "table: wgEncode...PkRep#", if you click the "describe table schema" button you will find definitions for signalValue and peak. Scrolling down you will find the related Track Description for UW Histone with the explanation for peak calling under "Methods" and the laboratory contact under "Credits".

ENCODE SOFTWARE TOOLS

Question:
How do I learn more about peak calling algorithms used to generate narrowPeak and broadPeak files?

Response:
An excellent resource to review is the ENCODE Software Tools page, located on the lefthand side of the ENCODE portal under "Software Tools." Click through to the Software Tools Used to Create the ENCODE Resource and here you can find references for the various peak calling algorithms under "ChIP-seq Peak Callers".

By visiting various ENCODE tracks such as HAIB TFBS, SYDH TFBS, or UW Histone you can learn more about the processes each lab used to generate peaks, and pick a method suitable for your data. Since these data were not generated by the UCSC Browser group, questions about the data methods need to be directed to the corresponding lab. Under the "Credits" section you will find a contact for further questions left unanswered by reading the descriptions.

ENCODE FILES

Question:
What program reads ".bb" TFBS files from ENCODE? I am interested in looking at the AWG TFBS data. I downloaded the files and one is called: spp.optimal.wgEncodeBroadHistoneGm12878CtcfStdAlnRep0_VS_wgEncodeBroadHistoneGm12878ControlStdAlnRep0.bb

However, I do not have a program that can open this file. What is the program for this file and where can I find it?

Response:
Files ending in ".bb" are bigBed files. Click here for extensive information on the bigBed format and how to extract data with different binary utilities located in this directory.

HUB EXAMPLES

Question:
I am making a public hub for my paper, is there an example html file to use for my data description?

Response:
The browser's public hubs provide excellent examples of hub documentation. Here are two examples of track description pages from the ENCODE Analysis hub:

http://ftp.ebi.ac.uk/pub/databases/ensembl/encode/integration_data_jan2011/hg19/uniformTfbs.html
http://ftp.ebi.ac.uk/pub/databases/ensembl/encode/integration_data_jan2011/hg19/uniformRNA.html

Useful tips when writing your track descriptions:

It is best to assume a broad audience of students as well as researchers. Spelling out common acronynms, for example, may be useful for those who are new to genomics.

The paper's abstract may be a good start for your track's "Description" section.

Provide as much detail as possible in the "Methods" section.

A email address must be prominently displayed for questions relating to the track.

Other Examples:

Here are a few good examples of hub structure and configuration from the ENCODE Analysis hub:

http://ftp.ebi.ac.uk/pub/databases/ensembl/encode/integration_data_jan2011/hub.txt
http://ftp.ebi.ac.uk/pub/databases/ensembl/encode/integration_data_jan2011/genomes.txt
http://ftp.ebi.ac.uk/pub/databases/ensembl/encode/integration_data_jan2011/hg19/trackDb.txt

Note: We recommend a minimal number of default visible tracks in your trackDb.txt to quicken hub loading time and to avoid overwhelming users. For more suggestions on hub structure, please see our Public Hub Guidelines wikipage. Also, for help defining unfamiliar terms, you may want to see the Hub Track Database Definition's table of contents.

Updated 15 August 2013