BeadLevelNormalise         package:beadarray         R Documentation

_B_e_a_d _L_e_v_e_l _N_o_r_m_a_l_i_s_a_t_i_o_n

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

     A single function which combines the steps of reading bead level
     data, normalisation and summarising.

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

     BeadLevelNormalise(targets, imageManipulation="sharpen",
                  backgroundCorrect="none", probes=NULL,imagesPerArray=2)

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

 targets: A targets object defining which bead level files to read

imageManipulation: string specifying what image processing steps to use
          to get foreground intensities

backgroundCorrect: string specifying what background correction (if
          any) to do

  probes: an optional list of ProbeIDs to use summarise

imagesPerArray: integer to specify how many images comprise each array.
          Used when we have to strips per array on the Human-Ref6
          arrays

_D_e_t_a_i_l_s:

     This is a memory and space efficient method for reading and
     normalising a large number of arrays at the bead level and based
     on the work of Henrik Bengsson. It comprises of a pre-processing,
     normalisation and probe summary.

     Quality control should be performed prior to this function is
     order to identify any outlier arrays and decide upon an image
     processing strategy.

     In the first stage of the function, we read each array in turn
     using the image processing steps specified by the
     imageManipulation and backgroundCorrect parameters and keep a
     running total of the value of each quantile of the distribution.

     For Human-6 arrays where each array is in fact two images, we read
     each image separately and bind together the resulting intensities
     before computing quantiles.

     Each array is then read a second time, normalised to the target
     distribution and then summarised.

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

     ExpressionSetIllumina containing summarised data

_A_u_t_h_o_r(_s):

     Mark Dunning

