[note: minor bug report buried deep in this email]
> Now I'm just having some problems with the BASE1/BASE2 column names.
>
> At last google and the mailing list help me out:
> http://www.mail-archive.com/[email protected]/msg00725.html
>
> It looks like I just have to create an extra value column called ratio1_2
actually I need a "formula" exactly as the mail suggests.
But I'm still getting a problem:
Status Error:
/usr/local/base_dir/base2dev_plugins/base1plugins/basehclust: Missing
reporterId/assayData columns or ratio field
Job parameters
Assay distance metric Pearson
New cluster point center of mass
Gene distance metric Pearson
Plugin directory /home/maccallr/Analysis Hierarchical clustering
ratioColumn ratio1_2
section clustering settings
Bioassay set background subtracted medians
Plugin configuration parameters
Parameter version 7 (7)
execName clusteringscript
Plugin executables path
/usr/local/base_dir/base2dev_plugins/base1plugins
geneAverages true
jobParameters [snip: imported from working base1 instance - no changes made]
leaveStdin true
leaveStdout false
maxChannels 2
minChannels 2
serialFormat false
usedColumns reporterId
usedFields ratio1_2
versionNumber 1.1
Since Base1PluginExecuter uses BioAssaySetExporter.exportBaseFileMatrix to
create a BASEfile, I tried to do this manually and get
"columns externalId assayData"
rather than
"columns reporterId assayData"
which I guess the basehclust binary is expecting?
Here's a snippet of the BASEfile I manually exported. See how the ratio1_2
column is all present and correct (thanks to the formula).
BASEfile
section assays
count 12
columns id name Generation OrganismPart
annotationColumns Generation OrganismPart
%
263 carcass_g1-cy3_SR-cy5 1 carcass
264 carcass_g1-cy5_SR-cy3 1 carcass
265 carcass_g3-cy5_SR-cy3 3 carcass
266 head_g1-cy3_SR-cy5 1 head
267 head_g1-cy5_SR-cy3 1 head
268 head_g3-cy5_SR-cy3 3 head
269 midgut_g1-cy3_SR-cy5 1 midgut
270 midgut_g1-cy5_SR-cy3 1 midgut
271 midgut_g3-cy5_SR-cy3 3 midgut
272 ovaries_g1-cy3_SR-cy5 1 ovaries
273 ovaries_g1-cy5_SR-cy3 1 ovaries
274 ovaries_g3-cy5_SR-cy3 3 ovaries
section spots
channels 2
assayFields ratio1_2
columns externalId assayData
assays 263 264 265 266 267 268 269 270 271
272 273 274
count 21100
%
4A3B-AAB-A-06 5.7225670458895 3.1218635882757 3.0588992108785 0.2639156720298
4.0223276353137 6.675902576748 3.4137582490339 2.3933277183965 2.7949516801256
2.7989004774074 2.481474953736 2.9725981696501
4A3B-AAB-A-12 2.6034823374857 1.8699111228203 2.1851485479522
0.46444012021543 1.5953635012332 2.6418038900175 1.1084148689112
0.81114689236237 1.2586586018848 1.3185744749523 1.3753473626496
1.0045607051797
4A3B-AAB-C-06 2.1968544953519 1.9939638540184 ...
I then tried to create a "reporterId" formula in the same way, but this caused
an uncaught exception (in the tomcat logs) when I try to view the column in
experiment explorer:
12:05:43,237 ERROR [jsp]:253 - Servlet.service() for servlet jsp threw exception
java.lang.ClassCastException: java.lang.String cannot be cast to
java.lang.Number
at
net.sf.basedb.util.formatter.NumberFormatter.format(NumberFormatter.java:35)
at net.sf.basedb.clients.web.taglib.table.Cell.doEndTag(Cell.java:231)
at
org.apache.jsp.views.experiments.explorer.view.view_jsp._jspService(view_jsp.java:1450)
Not surprisingly a text column doesn't work too well in a formula.
It looks like I should recompile basehclust to look for externalId instead of
reporterId. This seems trivial having grepped the source code. Any thoughts?
cheers,
Bob.
>
> cheers,
> Bob.
>
> Johan Enell writes:
> > Hj Bob,
> >
> > The clustering from Base1 will not work in Base2. Or more precisely,
> > you will be able to run the plugin but you wont be able to se the
> > result. The visualization page in Base1 was a specific php page for
> > that plugin and we didn't want a solution like that in Base2.
> >
> >
> > When you configure Base1 plugins you should always start of from a
> > base1 plugin configuration file. From the plugin page in Base1 you
> > will find an export link with each plugin. Use that file to configure
> > the Base1PluginExecuter.
> >
> > /Johan
> >
> >
> > 11 sep 2007 kl. 16.58 skrev Bob MacCallum:
> >
> > >
> > > Hi,
> > >
> > > I was trying to find out if base1's hierarchical clustering would
> > > run under
> > > base2 - and got to the point where I need to provide parameter
> > > settings as
> > > XML. I could spend an hour figuring this out by hand (e.g. from
> > > the examples
> > > on the BASE2 demo server) but I'm convinced there's an easier way.
> > > I've
> > > searched the plugin site for documentation. Could someone help
> > > please?
> > >
> > > Since the hierarchical clustering isn't on the BASE2 demo server,
> > > does that
> > > mean it doesn't work?
> > >
> > > We know that there won't be any display tools available in BASE2
> > > for this, but
> > > we can work with the PNGs and assorted text files.
> > >
> > > many thanks,
> > > Bob.
> > >
> > >
> > >
> > > --
> > > Bob MacCallum | VectorBase Developer | Kafatos/Christophides Groups |
> > > Division of Cell and Molecular Biology | Imperial College London |
> > > Phone +442075941945 | Email [EMAIL PROTECTED]
> > >
> > > ----------------------------------------------------------------------
> > > ---
> > > This SF.net email is sponsored by: Microsoft
> > > Defy all challenges. Microsoft(R) Visual Studio 2005.
> > > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > > _______________________________________________
> > > The BASE general discussion mailing list
> > > [email protected]
> > > unsubscribe: send a mail with subject "unsubscribe" to
> > > [EMAIL PROTECTED]
> >
> >
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2005.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > _______________________________________________
> > The BASE general discussion mailing list
> > [email protected]
> > unsubscribe: send a mail with subject "unsubscribe" to
> > [EMAIL PROTECTED]
>
> --
> Bob MacCallum | VectorBase Developer | Kafatos/Christophides Groups |
> Division of Cell and Molecular Biology | Imperial College London |
> Phone +442075941945 | Email [EMAIL PROTECTED]
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> The BASE general discussion mailing list
> [email protected]
> unsubscribe: send a mail with subject "unsubscribe" to
> [EMAIL PROTECTED]
--
Bob MacCallum | VectorBase Developer | Kafatos/Christophides Groups |
Division of Cell and Molecular Biology | Imperial College London |
Phone +442075941945 | Email [EMAIL PROTECTED]
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
The BASE general discussion mailing list
[email protected]
unsubscribe: send a mail with subject "unsubscribe" to
[EMAIL PROTECTED]
