@@ -1,8 +1,8 @@

 Package: MotifDb

 Type: Package

 Title: An Annotated Collection of Protein-DNA Binding Sequence Motifs

-Version: 1.21.3

-Date: 2018-04-08

+Version: 1.21.4

+Date: 2018-05-10

 Author: Paul Shannon, Matt Richards

 Maintainer: Paul Shannon <pshannon@systemsbiology.org>

 Depends: R (>= 2.15.0), methods, BiocGenerics, S4Vectors, IRanges, Biostrings

@@ -4,6 +4,7 @@ exportMethods (

    export,

    show,

    query,

+   query2,

    motifToGene,

    geneToMotif,

    associateTranscriptionFactors

@@ -1,5 +1,5 @@

-setGeneric('query', signature='object', function(object, queryString, ignore.case=TRUE)

-              standardGeneric ('query'))

+setGeneric('query', signature='object', function(object, queryString, ignore.case=TRUE) standardGeneric ('query'))

+setGeneric('query2', signature='object', function(object, andStrings, orStrings=c(), notStrings=c(), ignore.case=TRUE) standardGeneric ('query2'))

 setGeneric('motifToGene', signature='object', function(object, motifs, source) standardGeneric('motifToGene'))

 setGeneric('geneToMotif', signature='object', function(object, geneSymbols, source, ignore.case=FALSE) standardGeneric('geneToMotif'))

 setGeneric('associateTranscriptionFactors', signature='object',

@@ -281,6 +281,45 @@ setMethod ('query', 'MotifList',

         object [indices]

       })

 #-------------------------------------------------------------------------------

+setMethod ('query2', 'MotifList',

+

+   function (object, andStrings, orStrings=c(), notStrings=c(), ignore.case=TRUE) {

+      find.indices <- function(queryString)

+         {unique(as.integer(unlist(sapply(colnames(mcols(object)),

+                 function(colname) grep(queryString, mcols(object)[,colname],ignore.case=ignore.case)))))

+             }

+          # setup defaults

+       and.indices <- list(seq_len(length(object)))

+       or.indices <- list(seq_len(length(object)))

+       not.indices <- list(c())

+

+       if(length(andStrings) > 0)

+          and.indices <- lapply(andStrings, find.indices)

+

+       if(length(orStrings) > 0)

+          or.indices <- lapply(orStrings, find.indices)

+

+       if(length(notStrings) > 0)

+          not.indices <- lapply(notStrings, find.indices)

+

+          # start with the indices of all elements

+       final.indices <- seq_len(length(object))

+

+         # get the cumulative intersection of all the "and" terms

+         # this steadily dimishes the set of indices

+       for(indices in and.indices)

+         final.indices <- intersect(final.indices, indices)

+

+         # lump all of the "or" terms together: they all get included

+       final.indices <- intersect(unlist(or.indices), final.indices)

+

+         # finally reduce the set to exclude all indices of all "not" terms

+       for(indices in not.indices)

+         final.indices <- setdiff(final.indices, indices)

+

+       object [final.indices]

+       })

+#-------------------------------------------------------------------------------

 # Addition on 2017/06/15 from Matt Richards

 # This will not exactly match JASPAR because units are PFM and JASPAR uses PCM

@@ -29,6 +29,7 @@ runTests = function ()

   test.subset ()

   test.subsetWithVariables ()

   test.query ()

+  test.query2()

   test.transformMatrixToMemeRepresentation ()

   test.matrixToMemeText ()

   test.export_memeFormatStdOut ()

@@ -438,6 +439,42 @@ test.query = function ()

 } # test.query

 #------------------------------------------------------------------------------------------------------------------------

+test.query2 <- function()

+{

+  print ('--- test.query2')

+  mdb = MotifDb

+

+  ors <- c("MA0511.1", "MA0057.1")

+  ands <- c("jaspar2018", "sapiens")

+  nots <- "cisbp"

+  x <- query2(mdb, andStrings=ands, orStrings=ors)

+  checkEquals(length(x), 2)

+  checkEquals(sort(names(x)),

+             c("Hsapiens-jaspar2018-MZF1(var.2)-MA0057.1", "Hsapiens-jaspar2018-RUNX2-MA0511.1"))

+

+  x <- query2(mdb, andStrings="MA0057.1")

+  checkEquals(length(x), 15)

+

+  x <- query2(mdb, andStrings=c("MA0057.1", "cisbp"))

+  checkEquals(length(x), 11)

+

+  x <- query2(mdb, andStrings=c("MA0057.1"), notStrings="cisbp")

+  checkEquals(length(x), 4)

+

+  x <- query2(mdb, andStrings=c("MA0057.1"), notStrings=c("cisbp", "JASPAR_2014"))

+  checkEquals(length(x), 3)

+

+  x <- query2(mdb, orStrings=c("mus", "sapiens"), andStrings="MA0057.1")

+  #checkEquals(sort(names(x)),

+

+    # do queries on dataSource counts match those from a contingency table?

+  sources.list = as.list (table (mcols(mdb)$dataSource))

+  checkEquals (length (query2 (mdb, 'flyfactorsurvey')), sources.list$FlyFactorSurvey)

+  checkEquals (length (query2 (mdb, 'uniprobe')), sources.list$UniPROBE)

+  checkEquals (length (query2 (mdb, 'UniPROBE')), sources.list$UniPROBE)

+

+} # test.query2

+#------------------------------------------------------------------------------------------------------------------------

 test.transformMatrixToMemeRepresentation = function ()

 {

   print ('--- test.transformMatrixToMemeRepresentation')