problem with hub visualization

[Clara BENOIT]

Hi,

I'm trying to visualize 2 bigwig files on UCSC.

I followed the step-by-step description of how to set up a track hub and I loaded it in UCSC (with no error message). The tracks appear in the browser but there is not any data inside... (see screenshot attached)

Here's my hub link : https://fasterdb.lyon.unicancer.fr/UCSC_hubs/sipp_hub/hub.txt.

I know that my bigwig files are correct because I can visualize them in IGV (see screenshot attached).

When I used hubcheck with the link above, I got this error :

Couldn't open /Users/clarabenoit/.hg.conf , No such file or directory

I used the following workaround : HGDB_CONF=/dev/null hubCheck https://fasterdb.lyon.unicancer.fr/UCSC_hubs/sipp_hub/hub.txt.

And I didn't get any error message.

Is there a problem with my hub that is not reported by hubcheck?

Thanks in advance,

Clara Benoit

[Brian Lee]

Dear Clara Benoit,

Thank you for using the UCSC Genome Browser and your question about the display of a hub bigWig file that is displaying at IGV. The issue stems from the chromosome names in the bigWig files being displayed. The files have 1 for chromosome 1 and MT for mitochondria, which are acceptable notations at IGV, whereas the UCSC browser needs the abbreviation "chr" in front, and uses M instead of MT.

Thank you for your input about hubCheck, we are planning to improve it and will definitely include reporting this kind of error. If you have more suggestions we would greatly enjoy hearing them!

Here are some steps I checked that you could use to change your annotation files so they will display at UCSC in a hub. If you do not have the original wiggle/bedGraph files, you will need to expand your bigWigs with the utility "bigWigToWig", then you would perform some command line changes to add"chr" to the front of each line and swap MT with M (and remove any comment lines). Lastly you would repackage the wig/bedGraph back up into a new binary bigWig with "wigToBigWig".

Here is the directory you would use to find the correct utility for your operating system by running "uname -a" on the command line, then download "bigWigToWig" and "wigToBigWig": http://hgdownload.soe.ucsc.edu/admin/exe/

To run the wigToBigWig function you will also need a file that outlines the sizes of each chromosome (chrom.sizes), you can acquire it by running this command:

curl -O http://genome.ucsc.edu/goldenPath/help/hg19.chrom.sizes

1. Expand your bigWig (if you don't have it already in the non-bigWig state).

bigWigToWig accepted_hits.bw new.wig

2. Change the MT to a M and remove comments and add chr to the start of every line:

cat new.wig | grep -v "#" | sed -e "s/MT/M/" | sed -e "s/^/chr/" > new.fixed.wig

3. Turn the fixed.wig, that now has chrM instead of MT and chr20 instead of 20, into a bigWig

wigToBigWig new.fixed.wig hg19.chrom.sizes new.fixed.bw

Now you can put the URL to that new.fixed.bw into your trackDb.txt, or even load it directly via custom tracks: http://genome.ucsc.edu/cgi-bin/hgTracks?hgt.customText=http://university.edu/lab/directory/new.fixed.bw

Thank you again for using the UCSC Genome Browser hubs feature and sharing the issues you are having. Please feel free to reply with more feedback or send further suggestions to our suggestion box at http://genome.ucsc.edu/cgi-bin/hgUserSuggestion.

If you have any further questions, please reply to gen...@soe.ucsc.edu. All messages sent to that address are archived on a publicly-accessible forum. If your question includes sensitive data, you may send it instead to genom...@soe.ucsc.edu.

All the best,

Brian Lee

UCSC Genome Bioinformatics Group

--