NAM (version 1.5.3)

SNP QC: SNP Quality Control

Description

A function for quality control. It may be used to count/remove neighbor repeated SNPs and markers with MAF lower than a given threshold. This function is also used for imputations.

Usage

snpQC(gen,psy=1,MAF=0.05,misThr=0.8,remove=TRUE,impute=FALSE)

Arguments

gen

Numeric matrix containing the genotypic data. A matrix with \(n\) rows of observations and (\(m\)) columns of molecular markers. SNPs must be coded as 0, 1, 2, for founder homozigous, heterozigous and reference homozigous. NA is allowed.

psy

Tolerance parameter for markers in Perfect SYymmetry (psy). This QC remove identical markers (aka. full LD) that carry the same information. Default is 1, which removes only SNPs 100% equal to its following neighbor.

MAF

Minor Allele Frequency. Default is 0.05. Useful to inform or remove markers below the MAF threshold. Markers with standard deviation below the MAF threshold will be also removed.

misThr

Missing value threshold. Default is 0.8, removing markers with more than 80 percent missing values.

remove

Logical. Remove SNPs due to PSY or MAF.

impute

If TRUE, impute missing values using Random Forest adapted from the package missForest (Stekhoven and Buhlmann 2012) as suggested by Rutkoski et al (2013).

Value

Returns the genomic matrix without missing values, redundancy or low MAF markers.

References

Rutkoski, J. E., Poland, J., Jannink, J. L., & Sorrells, M. E. (2013). Imputation of unordered markers and the impact on genomic selection accuracy. G3: Genes| Genomes| Genetics, 3(3), 427-439.

Stekhoven, D. J. and Buhlmann, P. 2012. MissForest - nonparametric missing value imputation for mixed-type data. Bioinformatics, 28(1), 112-118.

Examples

# NOT RUN {
data(tpod)
gen=reference(gen)
gen=snpQC(gen=gen,psy=1,MAF=0.05,remove=TRUE,impute=FALSE)
# }