HMMcopy (version 1.14.0)

HMMcopy-package: Bias-free copy number estimation and robust CNA detection in tumour samples from WGS HTS data

Description

HMMcopy is a package for making bias-free copy number estimation by correcting for GC-content and mappability bias in HTS readcounts. It also contains an implementation of the Hidden Markov Model to robustly segment a copy number profile into non-overlapping segments predicted to be of the same copy number state, and attributes a biological copy number aberration events to the segments.

Details

HMMcopy takes as input WIG format files generated by fast C++ tools distributed as part of the HMMcopy Suite, namely readcount, GC-content and mappability values for non-overlapping fixed width “bins” across the reference genome of interest. It then uses a filtering and LOESS model to correct the GC-content and mappability biases observed in the readcounts (Benjamini and Speed, 2012), and uses the corrected readcounts as a proxy of copy number. The resultant copy number profile is then segmented with a six state Hidden Markov Model, with a handful of quick visualization functions for quick viewing.

Package:

HMMcopy

Type:

Package

Version:

0.1.0

Date:

2011-09-06

License:

GPL-3

example("HMMcopy-package") for quick tour of functionality and visualization

vignette("HMMcopy") for detailed example

References

Yuval Benjamini and Terence P Speed. Summarizing and correcting the gc content bias in high-throughput sequencing. Nucleic Acids Res, 40(10):e72, May 2012.

Examples

# Read WIG file input
rfile <- system.file("extdata", "tumour.wig", package = "HMMcopy")
gfile <- system.file("extdata", "gc.wig", package = "HMMcopy")
mfile <- system.file("extdata", "map.wig", package = "HMMcopy")
uncorrected_reads <- wigsToRangedData(rfile, gfile, mfile)

# Correct reads into copy number
corrected_copy <- correctReadcount(uncorrected_reads)

# Segment copy number profile
segmented_copy <- HMMsegment(corrected_copy)

# Visualize one at a time
par(ask = TRUE)
plotBias(corrected_copy)
plotCorrection(corrected_copy)
plotSegments(corrected_copy, segmented_copy)