Array Data

The fi rst input fi le that is required is the transcriptional profi ling data. This
should be a tab-delimited text format fi le (which may conveniently be created
using the “Save As” feature from within Microsoft Excel or another spreadsheet
package). Each row of this fi le represents the gene expression vector for a
discrete spot on the arrays. The fi rst column should therefore be a text identifi er
for the spot (this will in most cases be a gene name). The user must ensure
that each row has a unique identifi er, or the behavior of the program will be
undefi ned. Additionally, there must be only one such column of identifi ers.
Future versions should support the inclusion of an optional description column,
but this is not supported at this time. Note that the order in which the gene
names appear in the fi rst column determines the order in which the genes will
be displayed on the screen. If you wish to use the program to display data
in genome order, this order must be refl ected in the order of the genes in
the input file. Note also that it is possible in principle to order genes in
any manner you wish, such as ordering genes by functional category. After
the identifi er column, the number of subsequent columns depends on the
number of measurements available for each array feature. Hence, each subsequent
column represents the log ratios from a particular array. It is crucial
that the ratios be log transformed (the base of the logarithm is unimportant) to
ensure that positive and negative ratios are symmetric around zero. DecCor2
will work with incomplete data sets (resulting, e.g., from spots that have
not passed some quality control measure), provided that empty spots in the
spreadsheet are fi lled in using the “dummy” value of –999 (again, conveniently
done with Excel). There must not be an initial header row describing the
experiments.
An example fi le format for array data is as follows (all columns are tabdelimited):
GeneA 2.13 0.51 1.07 0.25 –2.34
GeneB 1.63 –999 5.14 –0.38 1.4
GeneC –4.71 2.14 –999 3.23 –999