findHotspots: Find hotspots in a NucDyn object.
In nucleosome-dynamics/NucDyn: Nucleosome Dynamics

Description Usage Arguments Details Value Author(s) Examples

Find hotspots (groups of single-read changes) from a NucDyn object. This function helps in the analysis of the nucleosome dynamics by finding the regions with relevant changes in the individual position of the nucleosomes. There are 4 types of basic hotspots:

findHotspots(dyn, nuc, wins = 10000, indel.threshold = NULL,
  shift.threshold = NULL, indel.nreads = NULL, shift.nreads = NULL,
  mc.cores = 1)

## S4 method for signature 'NucDyn'
findHotspots(dyn, nuc, wins = 10000,
  indel.threshold = NULL, shift.threshold = NULL,
  indel.nreads = NULL, shift.nreads = NULL, mc.cores = 1)

`dyn`	NucDyn object with the dynamic to analyze.
`nuc`	list of two `GRanges` objects containing the nucleosome calls of experiment 1 and experiment 2.
`wins`	Size of the window in base-pairs where the relative scores are computed
`indel.threshold`	Maximum p-value for an `INCLUSION` or `EVICTION` hotspot to be considered significant.
`shift.threshold`	Maximum p-value for a `SHIFT +` or `SHIFT -` hotspot to be considered significant.
`indel.nreads`	Minimum number of reads in an `INCLUSION` or `EVICTION` hotspot.
`shift.nreads`	Minimum number of reads in a `SHIFT +` or `SHIFT -` hotspot.
`mc.cores`	If `parallel` support, the number of cores available. This option is only used if the provided sets are from more than one chromosome.

SHIFT +: Translational movement of nucleosomes, downstream (+).
SHIFT -: Translational movement of nucleosomes, upstream (-).
EVICTION: Nucleosome reads removed from that locus.
INCLUSION: Nucleosome reads added to that locus.

As translational and coverage changes can happen anywhere, only those involving a certain number of reads are reported. This number can by adjusted by the threshold parameter. If threshold is a character vector representing a percentage value (i.e., 60%), this will be automatically converted to the absolute value given by the corresponding percentile of the coverage in the window. If, instead, threshold is a numeric value, this value will be used as absolute threhold. It two adjacent hotspots with shifts in opposite directions are detected but one of them is relatively small in comparison with the other, but will be reported as shifts, disregarding the value of combined. We consider two hotspots of the same magnitude if the ratio between the number of reads in one and the other is smaller than same.magnitude. This ratio is always performed by using the larger number as numerator and the smaller as denominator; therefore, same.magnitude must always be greater of equal than 1. For example, with same.magnitude=2, we consider that 25 reads shifting downstream followed bby 17 reads shifting upstream will be of the same magnitude (25/17 == 1.47 < 2) and we will annotate it as a "DISPERSION". In another example, if we have 25 shifts downstream followed by only 5 shifts upstream (25/5 == 5 > 2), both hotspot will be annotated as "SHIFT".

A data.frame with the following columns:

start: initial position of the detected hotspot
end: final position of the detected hotspot
peak: point of largest change (shortest p-value)
nreads: Number of reads involved at peak position
score: average p-value in the hotspot weighted by the coverage of reads involved in the hotspot
type: The type of the hotspot (as listed above).
chr: Chromosome name.

Ricard Illa, Diana Buitrago diana.buitrago@irbbarcelona.org

    data(readsG2_chrII)
    data(readsM_chrII)
    data(nuc_chrII)
    dyn <- nucleosomeDynamics(setA=readsG2_chrII, setB=readsM_chrII)
    hs <- findHotspots(dyn, nuc_chrII)