tdhock/PeakSegPipeline: Genome-Wide Peak Prediction Using Supervised Learning and Optimal Segmentation

Implementations of supervised machine learning algorithms for peak detection in genomic data, as described in Hocking and Bourque (2020) <doi:10.1142/9789811215636_0033>. Functional Pruning Optimal Partitioning with up-down constraints, Hocking et al. (2018) <arXiv:1810.00117> is used for single-sample peak prediction (independently for each sample and genomic problem). A fast heuristic discrete segmentation algorithm, Hocking and Bourque (2015) <arXiv:1506.01286> is used for joint peak prediction (for each peak, jointly using all samples).

install.packages("remotes")
remotes::install_github("tdhock/PeakSegPipeline")