HTCexp-class: Class 'HTCexp'
In bioinfo-pf-curie/HiTC: High Throughput Chromosome Conformation Capture analysis

Description Details Objects from the Class Slots Methods Author(s) See Also Examples

A class for representing high throughput Chromosome Conformation Capture data from next-generation sequencing experiments.

The normPerExpected method estimates the expected interactions based on a the dependency on the genomic proximity between two loci. See getExpectedCounts function for details.

The normPerTrans method is based on the assumption that all trans contacts should be the same. Thus, the cis contacts can be normalized by the interaction level of trans data. The xtrans trans map has to share its 'xgi' ranges with the cis map, and the ytrans has to share its 'ygi' ranges with the cismap. The method is used to combine the normalization factor from x and y ranges. Must be ‘sum’, ‘mult’ or ‘mean’.

Objects can be created either by:

calls of the form new("HTCexp", intdata, GRanges, GRanges).
using the auxiliary function HTCexp and supplying contact Matrix with x and y intervals definition. The forceSymmetric option can used to force intra-chromosomal contact maps to be stored as symmetrical matrix.

intdata:: Dense or Sparse Matrix, holding the interaction level between each pairs of 'x-y' intervals. The 'y' intervals must be in rows, and the 'x' in columns.
ygi:: Genomic ranges of y intervals; see class granges for details
xgi:: Genomic ranges of x intervals; see class granges for details

c(x, ...): Combines 'x' and the signature("HTCexp") objects in '...' together. The results is an object of class signature("HTCList")
detail(x): signature("HTCexp"): a more detailed output of the experiment than provided by show.
divide(x): comparison of two signature("HTCexp") objects. Perform the division of the two contact matrices on the common 'x' and 'y' intervals. The operation is done only on the common intervals of both objects. If one of the two objects has a count to zero, the divided value will be NA
intdata(x): return the intdata Matrix counts. Note that triangular matrices are always returned as symmetric matrices.
export(x): Defunct. See exportC method
isBinned: return TRUE if the data are binned. The method tests if the 'x' and 'y' genome intervals are the same, if 90% of the bins have the same size and if the full genomic range is covered
forceSymmetric(x): force the interaction data to 'symmetricMatrix'
forceTriangular(x): force the interaction data to triangular, ie. symmetric. Lower triangle of the matrix is set to zero, therefore reducing the size of the data in memory
isIntraChrom(x): return TRUE if the current signature("HTCexp") object contains intrachromosomal contact data
isSymmetric(x): return TRUE if the contact map is symmetrical, i.e inherits the symmetricMatrix class
normPerReads(x): normalize the contact matrix by the total number of reads of the matrix.
normPerExpected(x, ...): normalize the contact matrix by the expected number of reads based on the distance between two loci. See details.
normPerZscore(x): Defunct. See normPerExpected method
normPerTrans(x, xtrans, ytrans, method="sum"): Normalize cis contact map based on the trans interactions. See details
plot(x): visualization method; Display an heatmap of the contact data. Refer to the documentation of mapC for more details of the plotting function
range(x): return the genomic range of the signature("HTCexp") object
seq_name(x): Defunct. See seqlevels method
seqlevels(x): return the sequence levels of the signature("HTCexp") object
show(x): summarized output of the experiment, with informations about the data dimension and the genomic region studied
substract(x): comparison of two signature("HTCexp") objects. Perform the substraction of the two contact matrices on the common 'x' and 'y' intervals. The operation is done only on the common intervals of both objects. If one of the two objects has a count to zero, the divided value will be NA
summary(x): return descriptive summary statistics about the contact map
x_intervals(x): return the xgi GRanges object defining the x intervals
y_intervals(x): return the ygi GRanges object defining the y intervals
xy_intervals(x): return both xgi and ygi objects as a GRangesList object

Nicolas Servant

GRanges-class,GRangesList-class,Matrix-class

data(Nora_5C)

## HTCexp descriptio
show(E14)
detail(E14)

## Is binned data ?
isBinned(E14$chrXchrX)

## Is a inter or intrachromsomal experiment ?
isIntraChrom(E14$chrXchrX)

## Bin the data
E14.bin <- binningC(E14$chrXchrX, binsize=100000, step=3)

## Divide by expected interaction counts
E14norm<-normPerExpected(E14.bin)

## Operation on HTCexp object
E14_d_MEF<-divide(normPerReads(E14$chrXchrX), normPerReads(MEF$chrXchrX))
E14_s_MEF<-substract(normPerReads(E14$chrXchrX), normPerReads(MEF$chrXchrX))

## Overlap with genomic annotation
require(rtracklayer)
gene <- import(file.path(system.file("extdata", package="HiTC"),"refseq_mm9_chrX_98831149_103425150.bed"), format="bed")
plot(E14$chrXchrX, tracks=list(RefSeqGene=gene))

## Not run: 
## normPerTrans data normalization applied on \href{http://genome.ucsc.edu/cgi-bin/hgFileUi?db=hg19&g=wgEncodeUmassDekker5C}{ENCODE data}.
ENCODE=import.my5C("./ENM-GM12878-R1.matrix")

## Look at raw contact map
mapC(ENCODE$chr7chr7)

## look at normalize by trans contact map
mapC(normPerTrans(ENCODE$chr7chr7, xtrans=ENCODE$chr7chr5, ytrans=ENCODE$chr5chr7))

## End(Not run)

## Not run: 
## Export
exportC(E14$chrXchrX, con="E14.csv")

## End(Not run)