Hello Alison,

I wanted to share with you the bitbucket ticket for the improvements to composite datatypes. As Ross described, this would be part of a larger project. The ability to save composite data using the floppydisc icon would be one example an impact of the proposed change.


https://bitbucket.org/galaxy/galaxy-central/issue/654/improvements-for-composite-datatypes

Thanks again for the feedback and use case example,

Best,

Jen
Galaxy team

On 7/19/11 8:44 PM, Ross wrote:
Hi, Alison

I've made some fixes to the Manhattan/QQ tool (galaxy-central revision
#dacf97a6c663) so it now handles your data - thanks again for
discovering these edge cases - the revised code is now available and
working on http://test.g2.bx.psu.edu/ where your data produces plots.
In particular X in your data, and for good measure Y and Mt are now
translated into 23,24 and 25 respectively so the code can now cope
with those inputs and they should appear in linear order in the
Manhattan x axis

The floppydisk link (download) problem remains to be fixed as it's
more complex. We only update the main server every few weeks so the
changes I just made will take a while to appear there. Meanwhile,
please try the revised tool by making yourself an account and then
uploading your data to http://test.g2.bx.psu.edu/ - please let me know
how you get on?

Of course, the (?FDR adjusted) p values in your data are a challenge -
typically, the Manhattan and QQ plots are run with unadjusted p values
which are expected to be distributed uniformly 0-1 under the null...
FYI, when I use the grouping tool on your data, I see that chromosomes
20,21 and 22 are absent from the file - is this expected? Group with a
count gives:

1       22910
10      9990
11      12764
12      13413
13      13068
14      11319
15      11175
16      8483
17      12017
18      10073
19      6927
2       18332
3       16173
4       15401
5       15259
6       15487
7       15217
8       12485
9       14560
X       3951

Thanks again for helping us improve Galaxy...

On Mon, Jul 18, 2011 at 10:52 PM, Alison Harrill<aharr...@thehamner.org>  wrote:
The first time I uploaded the file, I loaded in the -logP values
instead. After that, I changed the file to raw p values and reuploaded.
Do you think it would help to upload the file again, but change the
filename? I just wonder if that original file didn't get cleared out.

Alison

-----Original Message-----
From: Ross [mailto:ross.laza...@gmail.com]
Sent: Monday, July 18, 2011 8:51 AM
To: Alison Harrill
Subject: Re: [galaxy-bugs] Error message when trying to dowload
Manhattan plot using the web version

ok - thanks. These are not raw pvalues and your data is definitely
causing the Manhattan plot to fail.

This will take me a little time to figure out - I'll get back to you
in a day or two - thanks again for reporting this problem

On Mon, Jul 18, 2011 at 10:16 PM, Alison Harrill
<aharr...@thehamner.org>  wrote:
Good morning,

Sorry for taking a while to get back to you (I was out of town for the
weekend). I set permissions to share the dataset with you. Please let
me
know if you see an obvious workaround.

Thank you,
Alison


Alison Harrill, Ph.D.
Research Investigator
Institute for Drug Safety Sciences
Hamner Institutes for Health Sciences
phone: (919) 226-3142
fax: (919) 226-3150



-----Original Message-----
From: Ross [mailto:ross.laza...@gmail.com]
Sent: Sunday, July 17, 2011 7:25 AM
To: Alison Harrill
Subject: Re: [galaxy-bugs] Error message when trying to dowload
Manhattan plot using the web version

Can you please share a history with me (use this email address) so I
can take a look please?

I'm talking about links on the results page where you (should!) see
some graphs - when you click the history item display ('eye') icon -
not the history item save ('floppy disk') icon

On Sun, Jul 17, 2011 at 9:15 PM, Alison Harrill
<aharr...@thehamner.org>
wrote:
Hi, Ross,

Thanks so much for kindly following up with me. I tried the
workaround
you suggested, but it appears to not be working for me.
Right-clicking
prompts me to save a file called display.htm and when I open the
file,
it just reads "Unable to remove temporary library download archive
and
directory".

Best wishes,
Alison

-----Original Message-----
From: Ross [mailto:ross.laza...@gmail.com]
Sent: Sunday, July 17, 2011 4:09 AM
To: Alison Harrill
Subject: Re: [galaxy-bugs] Error message when trying to dowload
Manhattan plot using the web version

Hi, Alison - I wanted to be sure you are able to get your images from
the tool - the proper fix is going to take a while so I hope those
links are working for you?

cheers...


On Sat, Jul 16, 2011 at 9:39 AM, Ross<ross.laza...@gmail.com>  wrote:
Hi, Alison,

Thanks for reporting this error - I just confirmed that you have
found
a bug - the download function is currently not working correctly for
composite datatypes in the history.

Fortunately, the images themselves have download links under each
one
and can be saved - eg on a PC or linux workstation, while hovering
over the '(Click to download image Armitagep_manhattan.png)' press
the
right side mouse button and choose "save link as.." - you will be
able
to choose a location to save the image file on your local
workstation.
I hope this "work around" allows you to get on with your work -
sorry
for the inconvenience and thanks again for finding and reporting
this
error - we will fix it as soon as we can.



On Sat, Jul 16, 2011 at 6:46 AM, Alison Harrill
<aharr...@thehamner.org>  wrote:
Greetings,

I have a question about generating a Manhattan/QQ plot using the
web
tool. I
was able to successfully run the plot (the result box is green),
but
when I
try to download the results, I get a message that says "Unable to
remove
temporary library download archive and directory." Is there another
way to
retrieve the data?

Thank you,


___________________________________________________________
Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:

   http://lists.bx.psu.edu/

--
Jennifer Jackson
http://usegalaxy.org
http://galaxyproject.org/Support
___________________________________________________________
Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:

 http://lists.bx.psu.edu/

Reply via email to