cleaning_filter: Filter CNVs calls based on several parameters

In SinomeM/CNVgears: A Framework to Combine, Analize and Interpret CNVs Calling Results

Description

cleaning_filter "clean" a CNV calls dataset based on measures such as length, number of calls per sample and more.

Usage

cleaning_filter(
  results,
  min_len = 10000,
  min_NP = 10,
  blacklists = NULL,
  blacklist_samples = NULL,
  blacklist_chrs = NULL
)

Arguments

results,	a data.table, the output of read_results.
min_len,	minimum CNVs length, any shorter event will be filtered out. Default is 10000.
min_NP,	minimum CNVs points, any shorter event will be filtered out. Default is 10.
blacklists,	blacklist, a list of data.table or data.frame containing at least the following columns: "chr", "start", "end". Any event in a blacklists' region will be filtered out.
blacklist_samples,	character vector containing samples ID to filter out.
blacklist_chrs,	character vector containing chromosomes names in the package format, 1:22 for autosomes and 23 24 for chr X and Y.

Details

This function can be used together with summary in order to clean the dataset from possible noise and unwanted calls. It is generally recommended to briefly explore the data using summary and then proceeding to filter out any unwanted group of events. Mandatory arguments of the function are "results" and "min_len"/"min_NP", default values are the authors suggested minimal filtering step, however its quite common to filter anything shorter than 10 or even 50 kb, and/or any call made by less than 10 points. The use of blacklist of any kind is optional and should be done with caution, as it can filter potential biologically relevant events. Over-segmented samples the user wishes to exclude can be specified via blacklist_samples. Immunoglobulin regions can be generated with the function immuno_regions, while telomeric and/or centromeric regions can be obtained with telom_centrom.

Value

the CNVresults object results after the selected filters have been applied.

Examples

DT <- cleaning_filter(penn_22, min_len = 50000)

SinomeM/CNVgears documentation built on Nov. 21, 2021, 5:34 a.m.