mainAnalysis            package:RNAither            R Documentation

_W_r_a_p_p_e_r _f_u_n_c_t_i_o_n _f_o_r _f_u_l_l _a_u_t_o_m_a_t_e_d _a_n_a_l_y_s_i_s

_D_e_s_c_r_i_p_t_i_o_n:

     Performs a standard analysis of the data (quality and statistics)
     from a dataset file.

_U_s_a_g_e:

     mainAnalysis(header, dataset, flagForSameExp, listOfNormalizations, listOfArgs4norm, 
     listOfStatTests, listOfArgs4stat, multTestAdj, hitScoringVec1, hitScoringVec2, 
     posNegFlag, flag4Gsea, vecOfChannels, whichOnto)

_A_r_g_u_m_e_n_t_s:

  header: the header of a dataset file generated with
          'generateDatasetFile' 

 dataset: an R data frame generated with 'generateDatasetFile' 

flagForSameExp: either 0 or 1. If 1, all experiments defined in the
          column ' ScreenNb' in the dataset file must have the same
          design (same type and same number of replicates - exact plate
          layout is irrelevant) so that Spearman's correlation
          coefficient can be computed between experiments (each with
          summarized replicates) 

listOfNormalizations: a list of the normalization function to apply.
          Can be 'LiWongRank', 'varAdjust', 'divNorm',
          'quantileNormalization', 'BScore', 'ZScore',
          'ZScorePerScreen', 'subtractBackground', 'lowessNorm',
          'controlNorm' 

listOfArgs4norm: a list containing, for each element of '
          listofnormalizations', the arguments to be passed on 

listOfStatTests: a list of the statistical tests to perform. Can be
          'Ttest', 'MannWhitney', 'RankProduct' 

listOfArgs4stat: a list containing, for each element of '
          listofstattests', the arguments to be passed on 

multTestAdj: indicates the p-value correction for multiple testing -
          one of ' "holm"', ' "hochberg"', ' "hommel"', '
          "bonferroni"', ' "BH"', ' "BY"', ' "fdr"', or ' "none"' (Type
          '?p.adjust' for details))

hitScoringVec1: a vector of length 3 indicating (in that order): 

          - scoring according to p-value (0: no, 1: yes)

          - scoring according to ZScore with ZScore < threshold (0: no,
          1: yes), or according to ZScore < threshold and p-value <
          hitScoringVec2[1] (2)

          - scoring according to ZScore with ZScore > threshold (0: no,
          1: yes), or according to ZScore > threshold and p-value <
          hitScoringVec2[1] (2). 

          If hitScoringVec1[2] or hitScoringVec1[3] are equal to 2,
          hitScoringVec1[1] must be equal to one, otherwise p-values
          will not be computed. 

hitScoringVec2: a vector of length 3 indicating the thresholds for
          hitscoringvec1 

posNegFlag: either 0 (no controls available) or 1 (controls available) 

flag4Gsea: Can be:

          - either 0: No GSEA analysis is performed

          - or 1: A GSEA analysis is performed for each hit scoring
          method

          - or a binary vector that allows to choose which hit scoring
          method(s) will be used for a GSEA analysis. Hit scoring
          methods are sorted as follows: first, hits are scored
          according to the p-values of each test specified in '
          listOfStatTests '. Then, if the option of scoring hits
          according to p-values and Intensities is chosen (see '
          hitScoringVec1 ', for each test, a hit vector is generated.
          Finally, if the option of scoring hits according to
          Intensities only is chosen, hit vectors are generated for
          this option. 

vecOfChannels: a character vector containing the names of the channels
          to be used for quality plots, for example ' "SigIntensity"'
          or ' "NbCells"' 

whichOnto: one of the three GO hierarchies: ' "biological_process" ', '
          "molecular_function" ' or ' "cellular_component" ' - used for
          the GSEA analysis 

_V_a_l_u_e:

     Generates the html output files ' index.html' and '
     indexnorm.html' containing the quality analysis of raw and
     normalized data, respectively, and ' stats.html', containing the
     statistical analysis. If several normalization methods are
     applied, an 'indexnorm' file is generated after each.

_E_x_a_m_p_l_e_s:

     data(exampleHeader, package="RNAither")
     data(exampleDataset, package="RNAither")

     mainAnalysis(header, dataset, 0, list(controlNorm), list(list(1, 0, "SigIntensity", 1)), 
     list(Ttest, MannWhitney), list(list("l", 1, "SigIntensity", "GeneName"), 
     list("l", 1, "SigIntensity", "GeneName")), "none", c(1, 0, 0), c(0.05, 0, 0), 1, 
     1, c("SigIntensity", "NbCells"), "biological_process")

