xRA: Short/Long Recurrent Abnormalities detection
In cghRA: Array CGH Data Analysis and Visualization

Description Usage Arguments Value Note Author(s) References See Also

These functions extract Short Reccurent Abnormalities (SRA) and Long Reccurent Abnormalities (LRA) from a CGH array series, as described by Lenz et al. (2008).

The processing core xRA is common for both analysis, but is not intended to be called directly. Use the SRA and LRA wrappers instead.

  xRA(segTables, value = "copies", states = list(deletion=c(-Inf,-0.5), gain=c(0.5,Inf)),
    sampleMin = 2, quiet = FALSE, lengthMax, lengthMin, gaps.width, gaps.ratio)
  SRA(...)
  LRA(...)

`segTables`	An eventually named `list` of `data.frame`s to reshape. All the `data.frame`s must contain at least "chrom" (character), "start" (integer), "end" (integer) columns, and the column defined by `value`. Can also be a single `data.frame` containing all the segments, with a `.sampleIdentity` integer column.
`value`	Single character value, the column name from which extract values that will fill the output cells.
`states`	A named `list` of numerics defining the boundaries of each state. Each state may be defined by a single value (the only value in `segParallel` to link to the state) or by two boundaries (the lower boundary is part of the state, the upper one is not). `Inf` and `-Inf` can be used as boundaries.
`sampleMin`	Single numeric value, minimal amount of samples in the 'overlapping group'. If lesser than 1, interpreted as a proportion of the sample count. Large values decrease processing time and SRA amounts.
`quiet`	Single logical value, whether to print diagnostic `message`s or not.
`lengthMax`	Single integer value, segments larger than this value will be filtered out (25 Mb for SRA, `NA` for LRA). Use `NA` to disabled length filtering.
`lengthMin`	Single integer value, segments shorter than this value will be filtered out (`NA` for SRA, 15 Mb for LRA). Use `NA` to disabled length filtering.
`gaps.width`	Single integer value, alterated segments separated by a gap shorter than this value will be merged (see also 'gaps.ratio'; 500 kb for SRA, 10 Mb for LRA). Use `NA` to disabled gap filling.
`gaps.ratio`	Single numeric value, for a gap to be filled its two neighbors must be this value larger than it (see also 'gaps.width'; 1 for SRA, 1.5 for LRA). Use `NA` to disabled gap filling.
`...`	The `SRA` and `LRA` functions are only wrappers to `xRA` with distinct `lengthMax`, `lengthMin`, `gaps.width` and `gaps.ratio` values, all other arguments are passed through to `xRA`.

Returns a list with a data.frame for each state :

`chrom`	Chromosomal location.
`inPeak`	Numeric, proportion of the sample series in the 'overlapping group'.
`overlap.start, overlap.end`	Integer, position on the chromosome for the highest peak of the SRA (region covered by the whole 'overlapping group').
`start, end`	Integer, position on the chromosome for the SRA itself (largest region covered by 2/3 of the 'overlapping group').
`extended.start, extended.end`	Integer, position on the chromosome for the extended SRA (largest region covered by 1/3 of the 'overlapping group').

For Long Reccurent Abnormalities, Lenz et al. suggest to filter out regions involved in abnormal chromosome arms. For technical reasons, this filter was NOT implemented.

Sylvain Mareschal

Lenz G et al. "Molecular subtypes of diffuse large B-cell lymphoma arise by distinct genetic pathways". Proc Natl Acad Sci U S A. 2008 Sep 9;105(36):13520-5 (Supporting Information)

STEPS

cghRA documentation built on May 2, 2019, 3:34 a.m.