NarrowPeaks (version 1.12.0)
Shape-based Analysis of Variation in ChIP-Seq using Functional PCA
Description
The package applies a functional
version of principal component analysis (FPCA) to: (1) Process
data in wiggle track format (WIG) commonly produced by ChIP-Seq
peak callers by applying FPCA over a set of read-enriched
regions (ChIP-Seq peaks). This is done in order to shorten the genomic
locations accounting for a given proportion of variation among
the enrichment-score profiles. The function 'narrowpeaks'
allows splitting and trimming binding sites in
close proximity to each other, narrowing down the length of
the putative transcription factor binding sites while
preserving the information present in the variability of the
dataset and capturing major sources of variation. (2) Analyse
differential variation between multiple ChIP-Seq samples with
replicates. The function 'narrowpeaksDiff' quantifies differences
between the shapes, and uses Hotelling's T2 tests on
the functional principal component scores to identify
significant differences across conditions.