View source for Making a hub for a cell browser

This page is divided into two sections:

# A discussion of makeCbHub and it's options
# How to make a [https://genome.ucsc.edu/goldenPath/help/hgTrackHubHelp.html#Intro track hub] for a cell browser utilizing makeCbHub

= <code>makeCbHub</code> = 

makeCbHub will generate the trackDb stanzas for a set of tracks that you can then use to create a track hub. This section will go through the various options for the script and how to use them 

== Required option: <code>fileDir</code> == 

There is one required option: <code>fileDir</code>. This is a directory of big* files or a directory of directories containing big* files. 

For example, use the commands below to generate the trackDb stanzas for a single composite track in the mouse-brain-cutandtag dataset:
<pre>
cd /hive/data/inside/cells/datasets/mouse-brain-cutandtag/hub
makeCbHub olig2

Output:
track olig2
compositeTrack on
shortLabel olig2
longLabel olig2
visibility dense
autoScale group
type bigWig

     track olig2_cluster_non_oligo
     parent olig2 on
     shortLabel cluster_non_oligo
     longLabel cluster_non_oligo
     type bigWig 0.000000 2358.294189
     autoScale group
     bigDataUrl olig2/cluster_non_oligo.bw
     visibility dense

     track olig2_cluster_oligo
     parent olig2 on
     shortLabel cluster_oligo
     longLabel cluster_oligo
     type bigWig 0.000000 280.645325
     autoScale group
     bigDataUrl olig2/cluster_oligo.bw
     visibility dense
</pre>

As you can see, it works, though it wouldn't be particularly pretty to look at the Genome Browser. The labels are not very human-friendly and both tracks will be colored the same, default color, black. 

==Options to customize output==

The six optional arguments for makeCbHub allow you greater control over what’s put into these trackDb stanzas, including shortLabels, longLabels, and colors.

===Composite track labels: <code>-d/--datasetList</code>===

The option <code>-d/--datasetList</code> serves two purposes:
# If <code>fileDir</code> contains multiple dirs, you can specify which of those you want to build trackDb stanzas for
# By default, the directory names under <code>fileDir</code> are used as the labels for the composite/parent tracks in the trackDb. However, these are required to be all lowercased (e.g. h3k27ac, bw, or clusters). This option allows one to specify the casing used for the short/long labels. 

<pre>
makeCbHub -d “Rad21 Olig2” bw/

track rad21
compositeTrack on
shortLabel Rad21
longLabel Rad21
visibility dense
autoScale group
type bigWig
...

track olig2
compositeTrack on
shortLabel Olig2
longLabel Olig2
visibility dense
autoScale group
type bigWig
...
</pre>

This command assumes that in bw/ (<code>fileDir</code>), there are two directories: rad21 and olig2, but it will use Rad21 and Olig2 as the shortLabel/longLabel for those composites in the trackDb. 

===Individual track labels: <code>-s/--shortLabel</code> and <code>-l/–longLabel</code>===

The options <code>-s/--shortLabel</code> and <code>-l/–longLabel</code> allow you to control the short and long labels of the individual tracks in the composites. By default the script uses the file names as the labels, which, depending on how the files are named, can be pretty messy:
<pre>
cd /hive/data/inside/cells/datasets/olg-eae-ms/eae-multiomics/hub
makeCbHub bw/
</pre>
Which results in:

     ...
     track bw_P21208_1004_OPC_Ctr_RND1_peaks
     parent bw on
     shortLabel P21208_1004_OPC_Ctr_RND1_peaks
     longLabel P21208_1004_OPC_Ctr_RND1_peaks
     type bigWig 0.000000 160.000000
     autoScale group
     bigDataUrl bw/P21208_1004_OPC_Ctr_RND1_peaks.bw
     visibility dense
     ...

However, if we rebuild the trackDb with these options:
<pre>
cd /hive/data/inside/cells/datasets/olg-eae-ms/eae-multiomics/hub
makeCbHub -s shortLabels.tsv -l acronyms.sorted.tsv bw/
</pre>
Which outputs:

     ...
     track bw_P21208_1004_OPC_Ctr_RND1_peaks
     parent bw on
     shortLabel OPC_Ctr
     longLabel OPC_Ctr - Control oligodendrocyte precursor cells
     type bigWig 0.000000 160.000000
     autoScale group
     bigDataUrl bw/P21208_1004_OPC_Ctr_RND1_peaks.bw
     visibility dense
     ...

The shortLabels file contains two columns.
# file name
# short label

Here's the line from the <code>shortLabels.tsv</code> used in the example above:

 P21208_1004_OPC_Ctr_RND1_peaks.bw       OPC_Ctr

The longLabels file follows the same format as the 'acronymFile' setting that can be specified in the cellbrowser.conf (and will mostly likely be the same file). The two columns are:
# short label
# long label

Here's the line from the <code>acronyms.sorted.tsv</code> used in the example above:

 OPC_Ctr Control oligodendrocyte precursor cells

'''Note''': the shortLabel/longLabel files can be csv or tsv format and their file names need to end with csv or tsv (e.g. shortLabels.tsv). 

===Colors: <code>-c/--colors</code> ===

If a track doesn't include the 'color' setting, the Genome Browser colors that track black by default. The <code>-c/–color</code> option allows you to specify the color for each track. The file follows the same format as the 'colors' setting that can be specified in the cellbrowser.conf (and will mostly likely be the same file). The two columns are 

# short label
# hexcode color (e.g. #834088), although RGB tuples (e.g. '131 64 136' or '131, 64, 136') are also accepted

=== Track description page: <code>-f/--html</code> ===

A track description page is there to briefly describe the data displayed by that track, any special display conventions, and contact info for the data creators. 

Creating and adding a track description page is discussed in more detail in the BLAH section. 

=== bigBed tracks: <code>-b/--bbType</code> ===

Most hubs you'll be building are going to be based on bigWig files. Occasionally, you may need to build one that includes a composite of bigBed files. The <code>-b/--bbType</code> option allows you to specify the bigBed type, e.g. 'bigBed 12 +', 'bigInteract', 'bigNarrowPeak'. The need for this option is because bigBed files are quite diverse. Many of the [http://genome.ucsc.edu/FAQ/FAQformat.html big* formats] except for bigWig essentially use the bigBed format as a base but use a different 'type' in the trackDb to tell the Genome Browser how to interpret and display the different columns. It would be difficult to auto-detect the type from just the info in the file itself. 

Here's an example using this option:
<pre>
cd /hive/data/inside/cells/datasets/human-enhancer-atlas/hub
makeCbHub --bbType bigNarrowPeak bb
</pre>
Which outputs:
<pre>
track bb
compositeTrack on
shortLabel bb
longLabel bb
visibility dense
autoScale group
type bigBed

     track bb_Acinar
     parent bb on
     shortLabel Acinar
     longLabel Acinar
     type bigNarrowPeak
     bigDataUrl bb/Acinar.bigBed
     visibility dense
</pre>



===Using all these options together===

Altogether, these options result in a solid track hub with user-friendly labels, colors, and a track description page. We'll skip the <code>-b/--bbType</code> since it's so rarely used. Here's a example you can run to see the results of combining these various options yourself: 
<pre>
cd /hive/data/inside/cells/datasets/olg-eae-ms/eae-multiomics/hub
makeCbHub -s shortLabels.tsv -l acronyms.sorted.tsv -c ../colors.csv -f track_desc.html bw/
</pre>
Which gets us:
<pre>
track bw
compositeTrack on
shortLabel bw
longLabel bw
visibility dense
autoScale group
type bigWig
html track_desc.html

     ...
     track bw_P21208_1004_OPC_Ctr_RND1_peaks
     parent bw on
     shortLabel OPC_Ctr
     longLabel OPC_Ctr - Control oligodendrocyte precursor cells
     type bigWig 0.000000 160.000000
     autoScale group
     color 131,64,136
     bigDataUrl bw/P21208_1004_OPC_Ctr_RND1_peaks.bw
     visibility dense
     ...
</pre>
In this case, you may want to adjust the short/long labels for the parent track from bw to something like "Chromatin Accessibility", but small improvements like that are relatively easy. You can take a look at the hub.txt used this example was drawn from: https://cells.ucsc.edu/olg-eae-ms/eae-multiomics/hub/hub.txt, or take a look at the hub in the [https://genome.ucsc.edu/s/mspeir/mm10_eae_hub Genome Browser]:

[[ File:Cellbrowser_gb_hub.png | x350px ]]

= Building a hub =

This section walks you through the process of creating a hub for a cell browser with tracks on a single assembly. If you are creating a hub with tracks on more than one assembly see the [http://genome.ucsc.edu/goldenPath/help/hubQuickStart.html Hub Quick Start] as it requires multiple configuration files.  

==File organization==

<code>makeCbHub</code> assumes a certain directory structure when it creates the trackDb stanzas. 

In your dataset directory, create a ‘hub’ directory where all of the hub-related files will live. In that hub directory, you will then create a directory for each [https://genome.ucsc.edu/goldenPath/help/trackDb/trackDbHub.html#compositeTrack composite/parent] track (directory names should be all lowercase): 
<pre>
cb_dataset_dir/
    |--> hub/
        |--> track_set_A/
            |--> track_A1.bw
            |--> track_A2.bw
            |--> etc…
        |--> track_set_B
            |--> track_B1.bw
            |--> etc…
</pre>

Dividing the individual tracks into composite/parent tracks will vary from dataset to dataset. For example, in the collection <code>mouse-brain-cutandtag</code>, individual tracks were divided into a composite track for each dataset in the collection (e.g. h3k27ac, h3k27me3, h3k27me3-cell-lines, h3k36me3, h3k4me3, olig2, rad21) as this was what was requested by the authors. In neuro-degen-atac, individual tracks were grouped according to their corresponding metadata field (e.g. broad-celltypes, clusters, neuronal-celltypes, neuronal-clusters). If you’re not sure how to group the tracks ask Max and/or the contributors. 

Finally, it’s best to make symlinks to the track files in the orig directory to prevent the unnecessary duplication of large amounts of files. See <code>human-enhancer-atlas/hub</code> and <code>fetal-chromatin-atlas/hub</code> as examples, where the bigWigs alone were 212 GB and 138 GB, respectively. (/hive has a ton of storage, but it's good to not waste space unnecessarily.)

==Track description page==

You'll need to create a track description page for the hub. A track description page describes the data that people are looking at, any display conventions to keep in mind for that data (e.g. coloring), a contact email, and a reference to the paper(s). If you keep the page fairly high-level, you can use the same one for all of the tracks in the hub. Even if there are multiple composite/parent tracks in your hub, you can use the same description page for all of them since often the data are very similar.

You can use the Genome Browser [https://genome.ucsc.edu/goldenPath/help/examples/hubExamples/templatePage.html template] as a starting point. Additionally, it can be useful to look at the track description pages for other hubs in the Cell Browser:

* <code>olg-eae-ms/eae-multiomics</code>: https://cells.ucsc.edu/olg-eae-ms/eae-multiomics/hub/track_desc.html
* <code>olg-eae-ms/eae-atac</code>: https://cells.ucsc.edu/olg-eae-ms/eae-atac/hub/track_desc.html
* <code>neuro-degen-atac</code> dataset: https://cells.ucsc.edu/neuro-degen-atac/track_desc.html (A good example of covering multiple composites with one page)

Once you have the page written, place it in the <code>hub/</code> directory.

==Setting up your <code>hub.txt</code>==

These lines will show up in any hub you create for the Cell Browser, though you have to adjust the values to be specific to your hub: 
<pre>
hub eae_multiomics
shortLabel scATAC-seq of EAE Mice
longLabel scATAC-seq of EAE Mice
genomesFile genomes.txt
email eneritz.agirre@ki.se
descriptionUrl https://www.sciencedirect.com/science/article/pii/S0896627321010898
useOneFile on

genome mm10

</pre>

Put these lines in a file called <code>hub.txt</code>.

Breakdown of what to do for each setting in the first stanza:
* <code>hub</code> - can be the same as the dataset short name (e.g. neuro-degen-atac) 
* <code>shortLabel</code> and <code>longLabel</code> - can be the same, though feel free to add more details to the longLabel if needed 
* <code>genomesFile</code> - will always be genomes.txt
* <code>email</code> - should be the email of the contributor
* <code>descriptionUrl</code> - can point to the paper or to the track description page you created
* <code>useOneFile</code> - will always be 'on'. The [http://genome.ucsc.edu/goldenPath/help/hgTracksHelp.html#UseOneFile useOneFile] option allows us to stick all of the hub configuration settings into one file.

After that, there should be a line break and then the <code>genome</code> setting. The value will be the name of the UCSC assembly the data was mapped to (e.g. hg38). You should confirm with the contributor the name of the genome assembly they used for their analysis. If they give you a non-UCSC name (e.g. GRCh38, GRCm39, etc.) you will need to map this to the UCSC assembly name. If it's a public dataset, you may be able to find this information in the methods section of the paper.

After that, you should add one more line break. This will ensure that the trackDb stanzas that you are going to add to the file in the next step will be properly separated from the <code>genome</code> line. 

== Running <code>makeCbHub</code>==

Once you have your big* files in place, it's time to use <code>makeCbHub</code> to create the trackDb stanzas. In addition to the one required argument, you'll likely want to use these three as well: 
# <code>-c/--colors</code>
# <code>-s/--shortLabels</code>
# <code>-f/--html</code>

Beyond those three, you may also end up using the <code>-d/--datasetList</code> and <code>-l/--longLabels</code> options as well if needed. 

Here's a set of example commands that you can use to create a set of trackDb stanzas using <code>makeCbHub</code>:
<pre>
cd /hive/data/inside/cells/datasets/olg-eae-ms/eae-multiomics/hub
makeCbHub -s shortLabels.tsv -l acronyms.sorted.tsv -c ../colors.csv -f track_desc.html bw
</pre>
Which outputs:
<pre>
track bw
compositeTrack on
shortLabel bw
longLabel bw
visibility dense
autoScale group
type bigWig
html track_desc.html

     track bw_P21208_1004_OPC_Ctr_RND1_peaks
     parent bw on
     shortLabel OPC_Ctr
     longLabel OPC_Ctr - Control oligodendrocyte precursor cells
     type bigWig 0.000000 160.000000
     autoScale group
     color 131,64,136
     bigDataUrl bw/P21208_1004_OPC_Ctr_RND1_peaks.bw
     visibility dense
     ...
</pre>

In reality, you'll be appending these lines to the end of your <code>hub.txt</code> file:
<code>
makeCbHub -s shortLabels.tsv -l acronyms.sorted.tsv -c ../colors.csv -f track_desc.html bw >> my_hub.txt
</code>

==Files in <code>/usr/local/apache/htdocs-cells</code> ==

Once you have the big* files, hub.txt, and track description page all ready, then you will need to make symlinks to the files in <code>/usr/local/apache/htdocs-cells</code>. 

First, navigate to your dataset's directory:

 cd /usr/local/apache/htdocs-cells/${your_dataset_dir}

Then, make a 'hub' directory:

 mkdir hub

Move into that directory, and create the proper symlinks:
<pre>
cd hub
ln -s /hive/data./inside/cells/datasets/${your_dataset_dir}/hub/hub.txt
ln -s /hive/data./inside/cells/datasets/${your_dataset_dir}/hub/track_desc.html
ln -s /hive/data./inside/cells/datasets/${your_dataset_dir}/hub/track_A
etc ...
</pre>
==<code>cellbrowser.conf</code> settings==

Now that you have everything in place in <code>/usr/local/apache/htdocs-cells</code>, you will need to add a few more settings to your <code>cellbrowser.conf</code> to have the links to the hub show up in the Cell Browser. 

In your <code>cellbrowser.conf</code>, add these lines:
<pre>
hubUrl="hub/hub.txt"
ucscDb="${ucsc_db_from_hub.txt}"
</pre>

The value for <code>ucscDb</code> should match the value from the <code>genome</code> line in your hub.txt.

Rebuild your hub using <code>cbBuild</code> to get the links to the Genome Browser to appear in the Cell Browser, as can be seen for the [https://cortex-dev.cells.ucsc.edu cortex-dev] dataset:

[[ File:Cellbrowser_gb_link.png | x250px ]]

[[ File:Cellbrowser_marker_gb_link.png | x250px ]]

As you push data from cells-test to cells-beta and to cells, the hub files should be automatically copied from server to server. 

[[Category:Cell Browser]]