The Modifier class — Modifier-class • RNAmodR

The Modifier class is a virtual class, which provides the central functionality to search for post-transcriptional RNA modification patterns in high throughput sequencing data.

Each subclass has to implement the following functions:

Slot nucleotide: Either "RNA" or "DNA". For conveniance the subclasses RNAModifier and DNAModifier are already available and can be inherited from.
Function aggregateData: used for specific data aggregation
Function findMod: used for specific search for modifications

Optionally the function settings<- can be implemented to store additional arguments, which the base class does not recognize.

Modifier objects are constructed centrally by calling Modifier() with a className matching the specific class to be constructed. This will trigger the immediate analysis, if find.mod is not set to FALSE.

Modifier(className, x, annotation, sequences, seqinfo, ...)

# S4 method for SequenceData
Modifier(
  className,
  x,
  annotation = NULL,
  sequences = NULL,
  seqinfo = NULL,
  ...
)

# S4 method for SequenceDataSet
Modifier(
  className,
  x,
  annotation = NULL,
  sequences = NULL,
  seqinfo = NULL,
  ...
)

# S4 method for SequenceDataList
Modifier(
  className,
  x,
  annotation = NULL,
  sequences = NULL,
  seqinfo = NULL,
  ...
)

# S4 method for character
Modifier(
  className,
  x,
  annotation = NULL,
  sequences = NULL,
  seqinfo = NULL,
  ...
)

# S4 method for list
Modifier(
  className,
  x,
  annotation = NULL,
  sequences = NULL,
  seqinfo = NULL,
  ...
)

# S4 method for BamFileList
Modifier(
  className,
  x,
  annotation = NULL,
  sequences = NULL,
  seqinfo = NULL,
  ...
)

Arguments

className

The name of the class which should be constructed.

x

the input which can be of the following types

SequenceData: a single SequenceData or a list containing only SequenceData objects. The input will just be used to file the data slot of the Modifier and must match the requirements of specific Modifier class.
BamFileList: a named BamFileList
character: a character vector, which must be coercible to a named BamFileList referencing existing bam files. Valid names are control and treated to define conditions and replicates

annotation

annotation data, which must match the information contained in the BAM files. This parameter is only required if x is not a SequenceData object or a list of SequenceData objects.

sequences

sequences matching the target sequences the reads were mapped onto. This must match the information contained in the BAM files. TThis parameter is only required if x is not a SequenceData object or a list of SequenceData objects.

seqinfo

An optional Seqinfo argument or character vector, which can be coerced to one, to subset the sequences to be analyzed on a per chromosome basis.

...

Additional otpional parameters:

find.mod: TRUE or FALSE: should the search for for modifications be triggered upon construction? If not the search can be started by calling the modify() function.
additional parameters depending on the specific Modifier class

All additional options must be named and will be passed to the settings function and onto the SequenceData objects, if x is not a SequenceData object or a list of SequenceData objects.

Value

a Modifier object of type className

Slots

nucleotide: a character value, which needs to contain "RNA" or "DNA"
mod: a character value, which needs to contain one or more elements from the alphabet of a ModRNAString or ModDNAString class.
score: the main score identifier used for visualizations
dataType: the class name(s) of the SequenceData class used
bamfiles: the input bam files as BamFileList
condition: conditions along the BamFileList: Either control or treated
replicate: replicate number along the BamFileList for each of the condition types.
data: The sequence data object: Either a SequenceData, SequenceDataSet or a SequenceDataList object, if more than one dataType is used.
aggregate: the aggregated data as a SplitDataFrameList
modifications: the found modifications as a GRanges object
settings: arguments used for the analysis as a list
aggregateValidForCurrentArguments: TRUE or FALSE whether the aggregate data was constructed with the current arguments
modificationsValidForCurrentArguments: TRUE or FALSE whether the modifications were found with the current arguments

Creation

Modifier objects can be created in two ways, either by providing a list of bamfiles or SequenceData/SequenceDataSet/SequenceDataList objects, which match the structure in dataType().

dataType() can be a character vector or a list of character vectors and depending on this the input files have to follow this structure:

a single character: a SequenceData is constructed/expected.
a character vector: a SequenceDataSet is constructed/expected.
a list of character vectors: a SequenceDataList is constructed/expected.

The cases for a SequenceData or SequenceDataSet are straight forward, since the input remains the same. The last case is special, since it is a hypothetical option, in which bam files from two or more different methods have to be combined to reliably detect a single modification (The elements of a SequenceDataList don't have to be created from the bamfiles, whereas from a SequenceDataSet they have to be).

For this example a list of character vectors is expected. Each element must be named according to the names of dataType() and contain a character vector for creating a SequenceData object.

All additional options must be named and will be passed to the settings function and onto the SequenceData objects, if x is not a SequenceData object or a list of SequenceData objects.