Statistical analysis failure?

[samp...@qut.edu.au]

We have just setup Corra, the practice dataset looks like it worked
using SuperHirn.. I recently ran some Q-tof data through the Corra
pipeline which consited of two LC runs under the same LC-MS
conditions. Feature Picking and Alignment look like they are working
(using SpecArray). However the Statistical analysis fails. I'm
wondering if this is due to my samples in particular or if it's
something else. Here is the corra.log information and below that is
the CorraStatistics.log information which looks like it gets halfway
through the process then stops:

---------------------------------------------------------------------------
COMMAND: ssh -x -i /data/Corra/conf/iproteom_id
ipro...@lyra.qut.edu.au qsub /home/iproteom/corra/users/course/OP5/
PBS/course_OP5_CSR.sh
---------------------------------------------------------------------------

INFO (Jun 14, 2011 7:30:43 AM)
org.systemsbiology.apps.corragui.server.executor.corrastats.TrackableCorraStatsExecutor
Job submitted: 868646
DEBUG (Jun 14, 2011 7:31:22 AM)
org.systemsbiology.apps.corragui.server.executor.corrastats.CorraStatsExecutor
Checking output of CorraStatistics.R
WARN (Jun 14, 2011 7:31:26 AM)
org.systemsbiology.apps.corragui.server.executor.corrastats.CorraStatsExecutor
File not generated: /hpcfs/iproteom/corra/users/course/OP5/OP5_hc.png
ERROR (Jun 14, 2011 7:31:26 AM)
org.systemsbiology.apps.corragui.server.executor.corrastats.CorraStatsExecutor
FAILED: Statistical Analysis course_OP5_CSR
ERROR (Jun 14, 2011 7:31:26 AM)
org.systemsbiology.apps.corragui.server.executor.corrastats.TrackableCorraStatsExecutor
checkOutput() failed on job: course_OP5_CSR


CorraStatistics.log info:
[1] "2011-06-14 07:30:47 Loading required libraries"
[1] "2011-06-14 07:30:50 Set up I/O"
OP5_pca.pdf
OP5_hc.pdf
OP5_volcano.pdf
OP5_volcano.tsv
[1] "2011-06-14 07:30:50 Read APML file 3"
[1] "2011-06-14 07:30:51 Beginning event-based parsing of /home/
iproteom/corra/users/course/OP5/aligned_features.apml"
[1] "Will return log2 transformed intensity values"
[1] "Starting document parse"
[1] "Found 2 total sources"
[1] "Dimensioning feature values: 4480 intensities 1792"
[1] "2011-06-14 07:30:51 aligned_features read"


Best wishes,
Dayle

[Micheleen Harris]

Hello Dayle,

I believe the problem is with the number of files you are using if I
am interpreting this correctly. Generally, you must have at least
duplicates of your conditions, although triplicates or more are best.
Corra statistics will not function well if there are not replicates.
If you have the data, go ahead and try this out. Or at least try out
adding more than two files and testing Corra statistics. Corra
statistics uses a pair-wise comparison, thus needing two or more
conditions.

Thanks for your email and continued interest in Corra. Feel free to
post any additional questions you might have.

Best,
Micheleen

> --
> You received this message because you are subscribed to the Google Groups "Corra" group.
> To post to this group, send email to co...@googlegroups.com.
> To unsubscribe from this group, send email to corra+un...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/corra?hl=en.
>
>