segmentation objects

Description

The segmentation class consists of matrices and lists. The components contain the output of the chromatin segmentation analysis. Loading the input data is optional. The object is returned as a result of calling learn_model or reading its already existing output.

Slots

model

list. The list consists of 6 items corresponding to the contents of the model_#.txt file. These are number_states and number_marks for the numbers of states and marks in the model; likelihood and probinit for the likelihood and the initial probabilities of the multi-state model; transitionprobs and emissionprobs for the probabilities of the transitions and emissions parameters of the model. Can be accessed using model.

emission

matrix. The matrix contains the emission parameters of n states (rows) for n marks (columns) corresponding to the contents of the emission_#.txt file. Can be accessed using emission.

transition

matrix. The matrix contains the transition parameters of n by n states corresponding to the contents of the transition_#.txt file. Can be accessed using transition.

overlap

list. A list of n number of cells/conditions items. Each item is a matrix of the overlap enrichment of n states (rows) at n genomic annotations (columns) corresponding to the contents of the <cell>_#_overlap.txt files. Can be accessed using overlap.

TSS

list. A list of n number of cells/conditions items. Each item is a matrix of the overlap enrichment of n states (rows) at n locations around the transcription start site (TSS) (columns) corresponding to the contents of the <cell>_#_TSS_neighborhood.txt files. Can be accessed using TSS.

TES

list. A list of n number of cells/conditions items. Each item is a matrix of the overlap enrichment of n states (rows) at n locations around the transcription end site (TES) (columns) corresponding to the contents of the <cell>_#_TES_neighborhood.txt files. Can be accessed using TES.

segment

list. A list of n number of cells/conditions items. Each item is a GRanges object containing the segmentation and assigned states as a metadata column 'state'. These contents correspond to the <cell>_#_segment.bed files. Annotations of the ranges are optional. Can be accessed using segment.

bins

list. A list of n number of cells/conditions items. Each item is a SummarizedExperiment object containing the binarized input data. The coordinates of the bins are saved as the rowRanges each assigned to a state and the binary data itself is saved as assay. Can be accessed using bins.

counts

list. A list of n number of cells/conditions items. Each item is a SummarizedExperiment object containing the read counts in bins. The coordinates of the bins are saved as the rowRanges each assigned to a state and the counts data itself is saved as assay. Can be accessed using counts.

See Also

learn_model