[Genome] genome graphs problem?

Ross Lazarus

unread,

Feb 8, 2009, 11:06:52 AM2/8/09

to gen...@soe.ucsc.edu, ross.l...@channing.harvard.edu

Help please! I've been using GG from Galaxy for about 2 years but today, when I
went to make some nice pictures for my grant renewal (!), there appears to be a
new problem.

Using a Galaxy constructed url like this:
http://genome.ucsc.edu/cgi-bin/hgGenome?db=hg18&hgGenome_dataSetName=Plink%20sample%20wga%20data%20case%20control%20_CaCo_gg.track&hgGenome_dataSetDescription=GalaxyGG_data&hgGenome_formatType=best%20guess&hgGenome_markerType=best%20guess&hgGenome_columnLabels=first%20row&hgGenome_maxVal=&hgGenome_labelVals=&hgGenome_maxGapToFill=25000000&hgGenome_uploadFile=http%3A%2F%2Fgalaxy.esphealth.org%2Fdisplay_as%3Fid%3D477%26display_app%3Ducsc&hgGenome_doSubmitUpload=submit

the data appear to upload fine, but the page that tells me how many snp and
where they mapped does not show the usual button to proceed to view the
karyotype, and when I view the page source, it is clearly truncated in the
middle of the form that usually shows that button - here's the source for the
UCSC page I'm getting after uploading (apparently successfully!) - note where it
ends - and no closing body or html tags:

====================source=========================================================
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>

<META HTTP-EQUIV="Content-Type" CONTENT="text/html;CHARSET=iso-8859-1">
<META http-equiv="Content-Script-Type" content="text/javascript">
<META HTTP-EQUIV="Pragma" CONTENT="no-cache">
<META HTTP-EQUIV="Expires" CONTENT="-1">
<TITLE>
Data Upload Complete (62 bytes) </TITLE>
<LINK REL="STYLESHEET" HREF="../style/HGStyle.css" TYPE="text/css">

</HEAD>
<BODY BGCOLOR="#FFF9D2" LINK="#0000CC" VLINK="#330066" ALINK="#6600FF"><A
NAME="TOP"></A>

<TABLE BORDER=0 CELLPADDING=0 CELLSPACING=0 WIDTH="100%">


<TR><TD COLSPAN=3 HEIGHT=40 >
<table bgcolor="#000000" cellpadding="1" cellspacing="1" width="100%%" height="27">
<tr bgcolor="#2636D1"><td valign="middle">
<table BORDER=0 CELLSPACING=0 CELLPADDING=0 bgcolor="#2636D1" height="24"><TR>
<TD VALIGN="middle"><font color="#89A1DE"> 

 <A HREF="../index.html?org=Human&db=hg18&hgsid=1605988" class="topbar">
Home</A>    
<A HREF="../cgi-bin/hgGateway?org=Human&db=hg18&hgsid=1605988"
class="topbar">
Genomes</A>    
<A
HREF="../cgi-bin/hgBlat?command=start&org=Human&db=hg18&hgsid=1605988"
class="topbar"> Blat</A>    
<A
HREF="../cgi-bin/hgTables?org=Human&db=hg18&hgsid=1605988&hgta_doMainPage=1"
class="topbar">
Tables</A>    
<A HREF="../cgi-bin/hgNear?org=Human&db=hg18&hgsid=1605988" class="topbar">
Gene Sorter</A>    
<A HREF="../cgi-bin/hgPcr?org=Human&db=hg18&hgsid=1605988" class="topbar">
PCR</A>    
<A
HREF="../cgi-bin/hgGenome?org=Human&db=hg18&hgsid=1605988&hgGenome_doPsOutput=on"
class="topbar">
PDF/PS</A>    
<A HREF="../cgi-bin/hgSession?org=Human&db=hg18&hgsid=1605988&hgS_doMainPage=1"
class="topbar">Session</A>   
<A HREF="../FAQ/" class="topbar">
FAQ</A>    

<A HREF="../goldenPath/help/hgGenomeHelp.html"
class="topbar">
Help</A>
 </font></TD>
</TR></TABLE>
</TD></TR></TABLE>
</TD></TR>


<TR><TD COLSPAN=3>

<TABLE WIDTH="100%" BGCOLOR="#888888" BORDER="0" CELLSPACING="0"
CELLPADDING="1"><TR><TD>
<TABLE BGCOLOR="#FFFEE8" WIDTH="100%" BORDER="0" CELLSPACING="0"
CELLPADDING="0"><TR><TD>
<TABLE BGCOLOR="#D9E4F8" BACKGROUND="../images/hr.gif" WIDTH="100%"><TR><TD>
<FONT SIZE="4"><b> 
Data Upload Complete (62 bytes)</b></FONT></TD></TR></TABLE>
<TABLE BGCOLOR="#FFFEE8" WIDTH="100%" CELLPADDING=0><TR><TH HEIGHT=10></TH></TR>
<TR><TD WIDTH=10> </TD><TD>

<FORM ACTION="../cgi-bin/hgGenome"><INPUT TYPE=HIDDEN NAME="hgsid"
VALUE="1605988">Loaded 15625346 elements from snp129 table for mapping.<BR>
Mapped 226064 of 228694 (98.8%) of markers<BR>

====================source ends here ===================================

--
Ross Lazarus, Associate Professor, DACP, Harvard Medical School.
Director of Bioinformatics, Channing Laboratory, BWH
181 Longwood Ave., Boston MA 02115, USA.

Jennifer Jackson

unread,

Feb 9, 2009, 4:08:11 PM2/9/09

to Ross Lazarus, gen...@soe.ucsc.edu

Hello,
I am unable to duplicate because that I do not have access to all the
data referenced in your link.
One thing to try is clearing your cart and starting the query/display over:
http://genome.ucsc.edu/cgi-bin/hgGateway
use the link here "Click here to reset the browser user interface
settings to their defaults."

If you still having problems with this page loading completely, please
let us know what browser/OS you are using. Also send any custom
tracks/data that would be required to duplicate the data (send directly
to me, not the mailing list). I will try the query and investigate the
problem.

Jennifer Jackson
UCSC Genome Bioinformatics Group

Jennifer Jackson

unread,

Feb 9, 2009, 8:43:03 PM2/9/09

to Ross Lazarus, gen...@soe.ucsc.edu

Hello Ross,

We have isolated your problem and are in the process of
implementing/testing the fix. Meanwhile, you are welcome to use/test
along with us at http://genome-test.cse.ucsc.edu/

Please let us know if you have any problems or observations,

Jennifer Jackson
UCSC Genome Bioinformatics Group

Ross Lazarus wrote:

Ross Lazarus

unread,

Feb 9, 2009, 9:05:31 PM2/9/09

to Jennifer Jackson, gen...@soe.ucsc.edu

Goodness me - thank you! That was quick! Yes, I'll try to test it as
soon as I can.

--
Ross Lazarus MBBS MPH
Associate Professor, Department of Ambulatory Care and Prevention
Director of Bioinformatics, Channing Laboratory

181 Longwood Ave., Boston MA 02115, USA.

Tel: +617 525 2730 Fax: +617 525 0958

Ross Lazarus

unread,

Feb 9, 2009, 9:16:59 PM2/9/09

to Jennifer Jackson, gen...@soe.ucsc.edu

Yes - it works there from a slightly faked Galaxy url so I think
you've nailed it thanks. Please let me know when you've rolled this
out to production?
Thanks again for sorting this out so efficiently.

Ross Lazarus

unread,

Jun 19, 2009, 4:38:18 PM6/19/09

to gen...@soe.ucsc.edu, ross.l...@channing.harvard.edu

It's back! Same problem that got fixed in February - after a successful
upload to genome graphs, the 'data upload complete' page is missing any
button to proceed to viewing the data!
Looks like a fix got overwritten?

Here's the source from today - even less at the end than the previous bug
report (see below for that previous report please) - note that the last
<FORM> element is not being closed - nor is there a </body></html>

================source of page 19 june
2009=======================================

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>

<META HTTP-EQUIV="Content-Type" CONTENT="text/html;CHARSET=iso-8859-1">
<META http-equiv="Content-Script-Type" content="text/javascript">
<META HTTP-EQUIV="Pragma" CONTENT="no-cache">
<META HTTP-EQUIV="Expires" CONTENT="-1">
<TITLE>

Data Upload Complete (67 bytes) </TITLE>

Home</A>    

<A HREF="../cgi-bin/hgGateway?org=Human&db=hg18&hgsid=135739082"

class="topbar">
Genomes</A>    

<A HREF="../cgi-bin/hgBlat?command=start&org=Human&db=hg18&hgsid=135739082"

class="topbar"> Blat</A>    

<A HREF="../cgi-bin/hgTables?org=Human&db=hg18&hgsid=135739082&hgta_doMainPage=1"

class="topbar">
Tables</A>    

<A HREF="../cgi-bin/hgNear?org=Human&db=hg18&hgsid=135739082"

class="topbar">
Gene Sorter</A>    

<A HREF="../cgi-bin/hgPcr?org=Human&db=hg18&hgsid=135739082"

class="topbar">
PCR</A>    

<A HREF="../cgi-bin/hgGenome?org=Human&db=hg18&hgsid=135739082&hgGenome_doPsOutput=on"

class="topbar">
PDF/PS</A>    

<A HREF="../cgi-bin/hgSession?org=Human&db=hg18&hgsid=135739082&hgS_doMainPage=1"

class="topbar">Session</A>   
<A HREF="../FAQ/" class="topbar">
FAQ</A>    

<A HREF="../goldenPath/help/hgGenomeHelp.html"
class="topbar">

Help</A>
 </font></TD>
</TR></TABLE>
</TD></TR></TABLE>
</TD></TR>


<TR><TD COLSPAN=3>

<TABLE WIDTH="100%" BGCOLOR="#888888" BORDER="0" CELLSPACING="0"
CELLPADDING="1"><TR><TD>
<TABLE BGCOLOR="#FFFEE8" WIDTH="100%" BORDER="0" CELLSPACING="0"
CELLPADDING="0"><TR><TD>
<TABLE BGCOLOR="#D9E4F8" BACKGROUND="../images/hr.gif" WIDTH="100%"><TR><TD>
<FONT SIZE="4"><b>

Data Upload Complete (67 bytes)</b></FONT></TD></TR></TABLE>

<TABLE BGCOLOR="#FFFEE8" WIDTH="100%" CELLPADDING=0><TR><TH
HEIGHT=10></TH></TR>
<TR><TD WIDTH=10> </TD><TD>

<FORM ACTION="../cgi-bin/hgGenome"><INPUT TYPE=HIDDEN NAME="hgsid"

VALUE="135739082">Loaded 15625346 elements from snp129 table for
mapping.<BR>

====================source ends here ============================

--
Ross Lazarus MBBS MPH

Associate Professor, Department of Population Medicine; Director of
Bioinformatics, Channing Laboratory

181 Longwood Ave., Boston MA 02115, USA.

Galt Barber

unread,

Jun 22, 2009, 6:01:31 PM6/22/09

to Ross Lazarus, gen...@soe.ucsc.edu

Hi, Ross.
There's nothing wrong.
You just got impatient and bailed out on the page
before it was done.

We did insert a flush(stdout) after the long snp markers
fetch finished (over 15 million now) so that the browser would
not get impatient waiting for the rest of the thing
to finish. However you yourself still need to be patient :)

Data Upload Complete (62 bytes)

Loaded 15625346 elements from snp129 table for mapping.

Mapped 226064 of 228694 (98.8%) of markers

These data are now available in the drop-down menus on the main page for
graphing.

FYI in GG upload format
you should have a marker followed by 1 or more numeric values that
are graphs. Poor GG is trying to interpret every (numeric) column
in your input after the snpId as another graph.
That's why you will see that 10 graphs have loaded.
Several of the columns will contain useless data.
The ones that say "log..." are the columns that
would probably be of most use to you.
Perhaps there is a way in Galaxy to select
just the columns you actually want to graph.

-Galt

> _______________________________________________
> Genome maillist - Gen...@soe.ucsc.edu
> https://lists.soe.ucsc.edu/mailman/listinfo/genome
>

Ross Lazarus

unread,

Jun 23, 2009, 9:45:57 AM6/23/09

to Galt Barber, gen...@soe.ucsc.edu

Thanks Galt - I did wait a very long time before bailing but you're right, I
did eventually bail...

Quick question: so why is GG trying to interpret anything other than the
first column as an identifier? I thought the logic for identifying marker
ids operated ONLY on that first column?
I've been leaving lots of columns that might be useful for other purposes as
the file has other uses - but you're right, the only ones of interest are
-log10p values - I guess I need to adjust the tools to produce a cut down gg
file as well as a more general tab delimited file including all the other
goodies....I don't want the user to have to do any manipulation - the whole
point of the GG interface is that it's one click - which roughly corresponds
to the upper limit of the attention span of my target users :)

Reply all

Reply to author

Forward