Heads up! This is a static archive of our support site. Please go to help.galaxyproject.org if you want to reach the Galaxy community. If you want to search this archive visit the Galaxy Hub search
Dear all:
I want to transform a text file to BedGraph format and show it on UCSC
genome browser, but it seems that in galaxy I cannot assign a column
that
has values in configuration screen (BedGraph format).
How should I do the above task in galaxy?
Thanks
John Wu
Hello John,
A file in bedgraph format should display at UCSC. I think that #3
(below) is your concern, but there are a few things to check if the
display at UCSC link is not showing up for the dataset:
1. File is tab delimited (not space or other) and four columns, such
as
(from the UCSC web site's example:
http://genome.ucsc.edu/goldenPath/help/bedgraph.html):
chr19 59302000 59302300 -1.0
chr19 59302300 59302600 -0.75
chr19 59302600 59302900 -0.50
chr19 59302900 59303200 -0.25
chr19 59303200 59303500 0.0
chr19 59303500 59303800 0.25
2. Use "Edit Attributes" to assign a database (click on pencil icon in
dataset). Confirm that this database is at UCSC by matching the
database
short name "dbkey" with that used in Galaxy. If the database is not at
UCSC, then Visualization within Galaxy is still an option (Trackster).
You can pull in other tracks from UCSC or other sources to view all of
the data together for custom genomes.
3. Use "Edit Attributes" to also assign the file format as "bedgraph".
The metadata will be automatically detected. The fourth column is not
explicitly labeled as being a data value for display, but that does
not
impact display. Not all values for all datatypes are labeled the same
as
external sites do, yet many display correctly based on the file format
type assigned.
4. Optional: create a track line using "Graph/Display Data -> Build
custom track for UCSC genome browser". This is not necessary, but does
allow for some user-defined labels. For full control over the track
and
(again, optional) browser lines used by UCSC, create this information
in
a text file on your computer, save, load into Galaxy, and merge with
the
bedgraph data using "Text Manipulation -> Concatenate datasets
tail-to-head".
If this does not help and you are still having issues on the public
Galaxy server at http://usegalaxy.org, please send a shared history
link
and I can take a look and provide feedback ("Options -> Share or
Publish, generate link, and email back directly to me). But, hopefully
the above helps to resolve anything that was unclear. I just tested a
bedgraph a few minutes ago, to confirm that there were no issue
between
Galaxy & UCSC for a sample bedgraph dataset, and it worked fine.
Best!
Jen
Galaxy team
--
Jennifer Jackson
http://usegalaxy.orghttp://galaxyproject.org/Support
Hi Jen,
Thanks for your answer.
However, when I followed your instructions, I found that the forth
column in my dataset, which is the expression level (or base depth
covered by short read sequences), would be treated as names instead of
values. Therefore I cannot see histogram on UCSC genome browser..
Could you help me more about the question? Thanks a lot!
Please see the following link to view my dataset.
http://main.g2.bx.psu.edu/u/johnwu/h/human-bodymap2
John Wu
To: 吳正華
Cc: galaxy-user@lists.bx.psu.edu
Subject: Re: [galaxy-user] BedGraph format
Hello John,
A file in bedgraph format should display at UCSC. I think that #3
(below) is your concern, but there are a few things to check if the
display at UCSC link is not showing up for the dataset:
1. File is tab delimited (not space or other) and four columns, such
as (from the UCSC web site's example:
http://genome.ucsc.edu/goldenPath/help/bedgraph.html):
chr19 59302000 59302300 -1.0
chr19 59302300 59302600 -0.75
chr19 59302600 59302900 -0.50
chr19 59302900 59303200 -0.25
chr19 59303200 59303500 0.0
chr19 59303500 59303800 0.25
2. Use "Edit Attributes" to assign a database (click on pencil icon in
dataset). Confirm that this database is at UCSC by matching the
database short name "dbkey" with that used in Galaxy. If the database
is not at UCSC, then Visualization within Galaxy is still an option
(Trackster).
You can pull in other tracks from UCSC or other sources to view all of
the data together for custom genomes.
3. Use "Edit Attributes" to also assign the file format as "bedgraph".
The metadata will be automatically detected. The fourth column is not
explicitly labeled as being a data value for display, but that does
not impact display. Not all values for all datatypes are labeled the
same as external sites do, yet many display correctly based on the
file format type assigned.
4. Optional: create a track line using "Graph/Display Data -> Build
custom track for UCSC genome browser". This is not necessary, but does
allow for some user-defined labels. For full control over the track
and (again, optional) browser lines used by UCSC, create this
information in a text file on your computer, save, load into Galaxy,
and merge with the bedgraph data using "Text Manipulation ->
Concatenate datasets tail-to-head".
If this does not help and you are still having issues on the public
Galaxy server at http://usegalaxy.org, please send a shared history
link and I can take a look and provide feedback ("Options -> Share or
Publish, generate link, and email back directly to me). But, hopefully
the above helps to resolve anything that was unclear. I just tested a
bedgraph a few minutes ago, to confirm that there were no issue
between Galaxy & UCSC for a sample bedgraph dataset, and it worked
fine.
Best!
Jen
Galaxy team
--
Jennifer Jackson
http://usegalaxy.orghttp://galaxyproject.org/Support
ADD COMMENT
• link
written
7.3 years ago by
吳正華 • 60
Hi,
I think you need to include a track line and specify "bedgraph", like
this
track type=bedGraph
Is that maybe the problem? It sounds like UCSC is treating your file
as a standard "bed", which it will do without the track line.
Jim
Hi Jim:
Thanks for your help!
I know how to make a bedgraph file, but it's just that I thought
galaxy
could automatically make a bedgraph file ( or track ) out from a text
file.
Maybe I was thinking wrong.
John
To: John Wu
Cc: galaxy-user@lists.bx.psu.edu
Subject: Re: [galaxy-user] BedGraph format
Hi,
I think you need to include a track line and specify "bedgraph", like
this
track type=bedGraph
Is that maybe the problem? It sounds like UCSC is treating your file
as a
standard "bed", which it will do without the track line.
Jim
in my dataset, which is the expression level (or base depth covered by
short
read sequences), would be treated as names instead of values.
Therefore I
cannot see histogram on UCSC genome browser..
display at UCSC link is not showing up for the dataset:
(from the UCSC web site's example:
dataset). Confirm that this database is at UCSC by matching the
database
short name "dbkey" with that used in Galaxy. If the database is not at
UCSC,
then Visualization within Galaxy is still an option (Trackster).
data together for custom genomes.
explicitly labeled as being a data value for display, but that does
not
impact display. Not all values for all datatypes are labeled the same
as
external sites do, yet many display correctly based on the file format
type
assigned.
track for UCSC genome browser". This is not necessary, but does allow
for
some user-defined labels. For full control over the track and (again,
optional) browser lines used by UCSC, create this information in a
text file
on your computer, save, load into Galaxy, and merge with the bedgraph
data
using "Text Manipulation -> Concatenate datasets tail-to-head".
server at http://usegalaxy.org, please send a shared history link and
I can
take a look and provide feedback ("Options -> Share or Publish,
generate
link, and email back directly to me). But, hopefully the above helps
to
resolve anything that was unclear. I just tested a bedgraph a few
minutes
ago, to confirm that there were no issue between Galaxy & UCSC for a
sample
bedgraph dataset, and it worked fine.
ADD REPLY
• link
written
7.3 years ago by
吳正華 • 60
