Hi there,

Looking at the output of the SplicingDiff files of CuffDiff, me and my
colleagues are preplexed about the output of the p_values and q_values.
We've tried different inputs of different samples to compare but never seem
to manage to get p_values smaller than 0.50 and we keep getting higher than
1 q_values (also smaller which we expect) which we think is strange too.

The input files we use for the CuffDiff are the CuffCompare of a combined
CuffCompare of a dataset, or the CuffCompare of just the two samples we want
to analyse. For the samples input files we use the TopHat files

Could you please help us get meaningful results for the SplicingDiff files
or help us understand the data?

The top 5 rows of our typical SplicingDiff file:
       test_id     gene_id gene                    locus sample_1 sample_2
1583  TSS11905 XLOC_028193    - chr5:134910259-134914719       q1       q2
2385  TSS12870 XLOC_030892    -   chr7:29976178-30008608       q1       q2
8005   TSS6887 XLOC_016656    -  chr18:47803031-47807892       q1       q2
10214  TSS9761 XLOC_022527    -  chr20:43128822-43138649       q1       q2
2818  TSS13383 XLOC_032450    - chr8:100899717-100905900       q1       q2
    status value_1 value_2    sqrt.JS. test_stat  p_value  q_value
1583      OK       0       0 0.000771867  0.797878 0.501645 164.5400
2385      OK       0       0 0.001548470  0.797809 0.505482  82.8991
8005      OK       0       0 0.003288510  0.797717 0.508184  55.5615
10214     OK       0       0 0.001414180  0.797620 0.510277  41.8427
2818      OK       0       0 0.007112780  0.797416 0.513678  33.6973

Thanks in advance for your most appreciated help,

Felix Mayr
The Galaxy User list should be used for the discussion of
Galaxy analysis and other features on the public server
at usegalaxy.org.  Please keep all replies on the list by
using "reply all" in your mail client.  For discussion of
local Galaxy instances and the Galaxy source code, please
use the Galaxy Development list:


To manage your subscriptions to this and other Galaxy lists,
please use the interface at:


Reply via email to