foldFilt. The general idea being that one might
want to create a Venn diagram showing probesets that are unique to
particular comparisons, or consistent between comparisons, and then might
want to output the probesets that are contained in each cell of the Venn
diagram.
vennSelectFC(eset, x, comps, order.by = "sum", method = "same", text = TRUE, html = TRUE, ...)ExpressionSet
object.foldFilt. See details for more
information.foldFilt See details for more
information.TRUETRUEprobes2table, in
particular, to change the argument to anncols which controls the
columns of hyperlinks to online databases (e.g., Entrez Gene, etc.). See
aaf.handler for more information.foldFilt. A small example
would be as follows:Assume an ExpressionSet exists that contains expression values for
three Affymetrix chips, say a control, and two experimentals. One might want
to know what probesets are different between each of the experimentals and
the control, and those that are different between both of the experimentals
and the control. We first make the comparisons, based on a fold change of 2
(or a difference of 1 on the log scale).
comps <- list(c(1,2), c(1,3))
This list indicates what comparisons we want. In this case 1vs2 and 1vs3.
out <- foldFilt(eset, fold = 1, groups = 1:3, comps = comps, compnames=c("Control vs experimental1", "Control vs experimental2"), save = TRUE)
By setting save = TRUE, we are saving a list, the first item being a vector
of the number of probesets in each comparison, the second item being an
indicator matrix showing up or down regulation based on a two-fold
difference. We could make a Venn diagram using this matrix with
vennCounts2 and vennDiagram. If we then
wanted to output the probesets in each cell of that Venn diagram, we could
use vennSelectFC as follows:
vennSelectFC(eset, out[[2]], comps)
One thing to note here is that the names of the resulting tables as well as
the columns containing the fold change values will be extracted from the
column names of the indicator matrix. This matrix will get its column names
from the 'compnames' argument to foldFilt, so it is best to use
reasonable names here. Also note that any character used in the 'compnames'
argument that is not a valid character for a file name will be stripped out.