UCSC Genome Browser tracks
The UCSC Genome Browser is an invaluable resource both for obtaining public sequencing data and for visualizing it.
Sometimes the UCSC Genome Browser at http://genome.ucsc.edu/ is pretty slow -- after all, it's a resource shared among the Eukaryotic genomics community.
But there's also a second "Beta test" version of the browser at http://hgwdev.cse.ucsc.edu/. It has slightly newer (and possibly less stable) code, but fewer people use it.
Configuring custom tracks
The UCSC Genome Browser has a "Custom Tracks" feature that lets you visualize your data using the Genome Browser web application. This data is visible only to you, not publically (unless you choose to share a link to it with others).
There are two approaches to visualizing your data in the UCSC Genome Browser:
- Directly upload a data file, in one of the supported formats.
- Your data is copied over the Internet to UCSC, where it is stored in tables and displayed as you browse.
- Appropriate for small to medium size files (up to a few MB).
- Host your data locally, and configure the UCSC Genome Browser with its URL.
- Your data resides in a location accessible via an HTTP or FTP public URL (e.g., our corral.tacc.utexas.edu:/gpfs/corral3/repl/utexas/BioinformaticsResource/web directory, that maps to http://).
- No data is copied to UCSC. You only tell the browser where to find the data when it is needed.
- Appropriate for large data sets (e.g. BAM files) that can be indexed for fast retrieval.
- Your data resides in a location accessible via an HTTP or FTP public URL (e.g., our corral.tacc.utexas.edu:/gpfs/corral3/repl/utexas/BioinformaticsResource/web directory, that maps to http://).
BED data
BED format is a simple 3 to 9 column format for location-oriented data.
See supported data formats for custom tracks for more information and examples.
VCF data
VCF data can only be configured as a URL, not uploaded directly. Directions are found at http://genome.ucsc.edu/goldenPath/help/vcf.html.
- The VCF file must be sorted by chromosome and position (most tools produce VCFs like this).
The VCF file must be compressed using bgzip:
module load tabix # also loads bgzip cd $BI/web bgzip progeria_ctcf.vcf
The VCF file must be indexed using tabix:
tabix -p vcf progeria_ctcf.vcf.gz
This has already been done, and the resulting files are at this URL: http://loving.corral.tacc.utexas.edu/bioiteam/ucsc_custom_tracks/, filename progeria_ctcf.vcf.gz. These are hg18 SNP calls from published Iyer Lab CTCF ChIP-seq data in Progeria cells. The VCF file was produce using Broad's GATK.
- Add custom tracks (be sure to pick assembly March 2006, NCBI36/hg18)
Here is the track configuration line
track type=vcfTabix name="progeria_ctcf_snp_calls" bigDataUrl="http://loving.corral.tacc.utexas.edu/bioiteam/ucsc_custom_tracks/progeria_ctcf.vcf.gz"
BAM data
BAM data can only be configured as a URL, not uploaded directly. Directions are found at http://genome.ucsc.edu/goldenPath/help/bam.html.
- The BAM file must be sorted and indexed using samtools. The .bam and .bai index file must reside in the same directory.
This has already been done, and the resulting files are at this URL: http://loving.corral.tacc.utexas.edu/bioiteam/ucsc_custom_tracks/, filename hela_totrna.sorted.bam. This is SE RNAseq data mapped directly to the human genome, hg19.
- Add custom tracks (be sure to pick assembly Feb 2009, NCBI37/hg19)
Here is the track configuration line
track type=bam name="hela_rnaseq" bigDataUrl="http://loving.corral.tacc.utexas.edu/bioiteam/ucsc_custom_tracks/hela_totrna.sorted.bam"
Here is another example, using paired end RNAseq data as processed using a tophat/cufflinks pipeline:
track type=bam name="rnaseq_bam" pairEndsByName=Y bigDataUrl="http://loving.corral.tacc.utexas.edu/bioiteam/ucsc_custom_tracks/accepted_hits.sorted.bam"
