panelcn.mops: Core copy number detection algorithm for targeted NGS panel...

In bioinf-jku/panelcn.mops: CNV detection tool for targeted NGS panel data

Core copy number detection algorithm for targeted NGS panel data

Description

This function performs the cn.mops algorithm for copy number detection in NGS data adjusted to targeted NGS panel data including the second quality control.

Usage

panelcn.mops(input, testi = 1, geneInd = NULL, classes = c("CN0", "CN1",
  "CN2", "CN3", "CN4"), I = c(0.025, 0.5, 1, 1.5, 2), priorImpact = 1,
  cyc = 20, normType = "quant", sizeFactor = "quant", qu = 0.25,
  quSizeFactor = 0.75, norm = 1, minReadCount = 5, maxControls = 25,
  corrThresh = 0.99, useMedian = FALSE, returnPosterior = TRUE)

Arguments

input	either an instance of "GRanges" or a raw data matrix, where columns are interpreted as samples and rows as genomic regions. An entry is the read count of a sample in the genomic region.
testi	positive integer that gives the index of the test sample in input. Default = 1
geneInd	vector of indices of rows input that are within target genes. These regions are not considered for chosing correlated reference samples. If NULL, all regions are considered for the correlation. Default = NULL
classes	vector of characters of the same length as the parameter vector "I". One vector element must be named "CN2". The names reflect the labels of the copy number classes. Default = c("CN0","CN1","CN2","CN3","CN4").
I	vector of positive real values containing the expected fold change of the copy number classes. Length of this vector must be equal to the length of the "classes" parameter vector. For human copy number polymorphisms the default is c(0.025,0.5,1,1.5,2).
priorImpact	positive real value that reflects how strong the prior assumption affects the result. The higher the value the more samples will be assumed to have copy number 2. Default = 1.
cyc	positive integer that sets the number of cycles for the algorithm. Usually after less than 15 cycles convergence is reached. Default = 20.
normType	type of the normalization technique. Each samples' read counts are scaled such that the total number of reads are comparable across samples. Options are "mean", "median", "poisson", "quant", and "mode". Default = "quant".
sizeFactor	parameter for calculating the size factors for normalization. Options are "mean", "median", "quant", and "mode". Default = "quant".
qu	Quantile of the normType if normType is set to "quant". Real value between 0 and 1. Default = 0.25.
quSizeFactor	Quantile of the sizeFactor if sizeFactor is set to "quant". 0.75 corresponds to "upper quartile normalization". Real value between 0 and 1. Default = 0.75.
norm	the normalization strategy to be used. If set to 0 the read counts are not normalized and cn.mops does not model different coverages. If set to 1 the read counts are normalized. If set to 2 the read counts are not normalized and cn.mops models different coverages. Default = 1.
minReadCount	if all samples are below this value the algorithm will return the prior knowledge. This prevents that the algorithm from being applied to segments with very low coverage. Default = 5.
maxControls	integer reflecting the maximal numbers of controls to use. If set to 0 all highly correlated controls are used. Default = 25
corrThresh	threshold for selecting highly correlated controls. Default = 0.99
useMedian	flag indicating whether "median" instead of "mean" of a segment should be used for the CNV call. Default = FALSE.
returnPosterior	flag that decides whether the posterior probabilities should be returned. The posterior probabilities have a dimension of samples times copy number states times genomic regions and therefore consume a lot of memory. Default = TRUE.

Value

an instance of "CNVDetectionResult".

Examples

data(panelcn.mops)
XandCB <- test
elementMetadata(XandCB) <- cbind(elementMetadata(XandCB), 
                                    elementMetadata(control))
result <- panelcn.mops(XandCB)

bioinf-jku/panelcn.mops documentation built on March 24, 2022, 1:19 a.m.