Jun 9, 2009, 12:14:07 AM (14 years ago)
Jari Häkkinen

Addresses #227

1 edited


  • plugins/base2/net.sf.basedb.illumina/trunk/README_PluginDetails

    r1111 r1113  
    1111This plug-in implements !BeadStudio like detection P-value calculations
    1212for Illumina expression data (see
    13 http://www.genomecenter.ucdavis.edu/expression_analysis/documents/GEX__normalization_070313.pdf
    1414on detection P-values.
    1616The plug-in will ''always'' base the detection P-value calculation on
    1717raw data values, ''i.e.'', the mean raw intensity for the different
    18 bead types. However, only reporters included in the bioassay set will
    19 be included in the distributions. The effect of the reporter filtering
    20 that may have occurred is that if no negative control bead types are
    21 left in the bioassay set, there is no distribution of negative
    22 controls to base the P-value detection calculation on.
     18signals. However, only reporters included in the bioassay set will be
     19included in the calculations. This means that if there are no negative
     20control bead types are left in the bioassay set, there is no
     21distribution of negative controls to base the P-value detection
     22calculation on. Or, if there are too few negative controls the
     23detection P-value calculation is unreliable. (There is a request,
     24http://baseplugins.thep.lu.se/ticket/204, for an user selectable
     25option to use all available negative controls available in raw data
     26irrespective the probe filtering state of the bio assayset.)
    24 Hence, typically you should use the detection P-value plug-in before
    25 you remove the negative controls from your analysis.
     28Hence, always use the detection P-value plug-in before removing
     29negative controls.
    27 There is a request, http://baseplugins.thep.lu.se/ticket/204, for an
    28 user selectable option to use all available negative controls in raw
    29 data even if no negative controls are left in a bioassay set.
     31The detection P-value plug-in does not filter the assays, it provides
     32the detection P-values usable in a filter step after running this
     35=== Parameters ===
     37The plug-in requires input of array type since detection P-values
     38depends on array type.
     40The user can also set a cut off parameter to exclude outliers within
     41the negative controls. The `cutoff` defines the acceptable negative
     42control signal range
     44median-MAD*cutoff < I < median+MAD*cutoff
     46where `MAD` is the median absolute deviation.
     49=== Implementation details ===
     51Each assay is treated separately, i.e., no samples are combined
     54Retrieve the raw expression value for all negative controls in the
     55current bioassay set.
     57For all signals `i` calculate the detection P-value as
     58`Pvalue = 1-i/N` where `i` is the rank of the signal relative to
     59the negative controls and `N` is the number of negative controls.
Note: See TracChangeset for help on using the changeset viewer.