Changeset 1122


Ignore:
Timestamp:
Jun 10, 2009, 11:41:55 PM (12 years ago)
Author:
Jari Häkkinen
Message:

Fixes #227.

Location:
plugins/base2/net.sf.basedb.illumina/trunk
Files:
2 edited

Legend:

Unmodified
Added
Removed
  • plugins/base2/net.sf.basedb.illumina/trunk/README_PluginDetails

    r1114 r1122  
    3535=== Parameters ===
    3636
    37 The plug-in requires input of array type since detection P-values
    38 depends on array type.
     37The plug-in requires input of array type since detection P-value are
     38calculated differently depending on array type.
    3939
    4040The user can also set a cut off parameter to exclude outliers within
     
    5050
    5151Each assay is treated separately, i.e., no samples are combined
    52 together.
     52together. All calculations are made on the raw bead-type level data,
     53i.e., on the average expression value for each bead type and raw data
     54is always used irrespective when in analysis the detection P-value is
     55calculated.
    5356
    54 Retrieve the raw expression value for all negative controls in the
    55 current bioassay set.
    56 
    57 Whole genome beadchips:
     57Pvalue calculation for whole genome arrays:
    5858
    5959For all signals `i` calculate the detection P-value as
    60 `Pvalue = 1-i/N` where `R` is the rank of the signal i relative to the
     60`Pvalue = 1-R/N` where `R` is the rank of the signal `i` relative to the
    6161negative controls and `N` is the number of negative controls.
    6262
    63 Others (DASL, miRNA, VeraCode DASL, and Focused Arrays):
     63Pvalue calculation for others array types (DASL, miRNA, VeraCode DASL,
     64and Focused Arrays):
    6465
    6566For all signals `i` calculate the detection P-value as
    6667`Pvalue = 1/2 - erf( [i-AvgControl]/StdControl/sqrt(2) )` where
    6768`AvgControl` is the average intensity of the negative controls,
    68 `StdControl` is the standard deviation of the ... (to be defined and
    69 implemented accordingly), and `erf` is the error function
     69`StdControl` is the standard deviation of the the negative controls,
     70and `erf` is the error function
    7071(http://mathworld.wolfram.com/Erf.html). The error function is used
    71 for arguments witin the range (-4,4). To save CPU cycles, the function
     72for arguments within the range (-4,4). To save CPU cycles, the function
    7273value for arguments outside this range is set to -1 and 1,
    7374respectively.
     
    7677----------------------------------------------------------------------
    7778{{{
    78 
    7979Copyright (C) 2009 Jari Häkkinen
    8080
  • plugins/base2/net.sf.basedb.illumina/trunk/src/net/sf/basedb/illumina/plugins/DetectionPValue.java

    r1121 r1122  
    9696    ( "Illumina detection P-value calculation",
    9797      "This plug-in implements BeadStudio like detection P-value calculations " +
    98       "for Illumina expression data. More details about this plug-in is found " +
    99       "at " + Illumina.URL + "\n\n Please send feedback to " + Illumina.EMAIL,
     98      "for Illumina expression data.\n\n" +
     99      "The plug-in will _always_ base the detection P-value calculation on " +
     100      "raw data values, i.e., the mean raw intensity for the different " +
     101      "signals. However, only reporters included in the bioassay set will be " +
     102      "included in the calculations. This means that if there are no negative " +
     103      "control bead types are left in the bioassay set, there is no " +
     104      "distribution of negative controls to base the P-value detection " +
     105      "calculation on. Or, if there are too few negative controls the " +
     106      "detection P-value calculation is unreliable. (There is a request, " +
     107      "http://baseplugins.thep.lu.se/ticket/204, for an user selectable " +
     108      "option to use all available negative controls available in raw data " +
     109      "irrespective the probe filtering state of the bio assayset.)\n\n" +
     110      "Hence, always use the detection P-value plug-in before removing " +
     111      "negative controls.\n\n" +
     112      "The detection P-value plug-in does not filter the assays, it provides " +
     113      "the detection P-values usable in a filter step after running this " +
     114      "plug-in.\n\n" +
     115      "Parameters\n" +
     116      "- The plug-in requires input of array type since detection P-value are " +
     117      "calculated differently depending on array type.\n" +
     118      "- The user can also set a cut off parameter to exclude outliers within " +
     119      "the negative controls. The `cutoff` defines the acceptable negative " +
     120      "control signal range " +
     121      "$median-MAD*cutoff < I < median+MAD*cutoff$ " +
     122      "where $MAD$ is the median absolute deviation.\n\n" +
     123      "Implementation details\n\n" +
     124      "Each assay is treated separately, i.e., no samples are combined " +
     125      "together. All calculations are made on the raw bead-type level data, " +
     126      "i.e., on the average expression value for each bead type and raw data " +
     127      "is always used irrespective when in analysis the detection P-value is " +
     128      "calculated.\n\n" +
     129      "Pvalue calculation for whole genome arrays:\n" +
     130      "For all signals $i$ calculate the detection P-value as " +
     131      "$Pvalue = 1-R/N$ where $R$ is the rank of the signal $i$ relative to the " +
     132      "negative controls and $N$ is the number of negative controls.\n\n" +
     133      "Pvalue calculation for others array types (DASL, miRNA, VeraCode DASL, " +
     134      "and Focused Arrays):\n" +
     135      "For all signals $i$ calculate the detection P-value as " +
     136      "$Pvalue = 1/2 - erf( [i-AvgControl]/StdControl/sqrt(2) )$ where " +
     137      "$AvgControl$ is the average intensity of the negative controls, " +
     138      "$StdControl$ is the standard deviation of the the negative controls, " +
     139      "and $erf$ is the error function " +
     140      "(http://mathworld.wolfram.com/Erf.html). The error function is used " +
     141      "for arguments within the range (-4,4). To save CPU cycles, the function " +
     142      "value for arguments outside this range is set to -1 and 1, " +
     143      "respectively.\n\n" +
     144      "More details about this plug-in is found at " + Illumina.URL +
     145      "\n\n Please send feedback to " + Illumina.EMAIL,
    100146      Illumina.VERSION,
    101147      Illumina.COPYRIGHT,
Note: See TracChangeset for help on using the changeset viewer.