monotone_regression {gpmap}

Perform monotone regression on a genotype-phenotype (GP) map

Description

The function uses partial_genotype_order and activeSet from the isotone package to do monotone regression (Leeuw et al., 2009) on a GP map.

Usage

monotone_regression(gpmap, plusallele)

Arguments

gpmap

An object of class gpmap

plusallele

An N vector of allele indexes (1 or 2)

Details

Element i in plusallele specifies the ordering of the genotypes at locus i, if the element is 1 then 11 < 12 < 22 and conversely if it is 2 then 22 < 12 < 11. monotone_regression calls partial_genotype_order to obtain the partial ordering of genotypic values for the given plusalleles. This partial ordering is then used together with the GP map itself as input to the activeSet function from the package isotone.

Value

monotone_regression returns the output from activeSet directly.

Author(s)

Arne B. Gjuvsland <[e-mail]> and Yunpeng Wang <[e-mail]>

References

Leeuw J, Hornik K and Mair P (2009) Isotone Optimization in R: Pool-Adjacent-Violators Algorithm (PAVA) and Active Set Methods. Journal of Statistical Software 32(5) [link]

Gjuvsland AB, Wang Y, Plahte E and Omholt SW (2013) Monotonicity is a key feature of genotype-phenotype maps. Front. Genet. 4:216. doi: 10.3389/fgene.2013.00216 [link]

Examples

data(GPmaps)

#Additive GP map is monotone 
monotone_regression(A,c(2,2))

#Pure AxA epistasis map
monotone_regression(AA,c(2,2))

#two-locus example in Cheverud & Routman (1995)
monotone_regression(mouseweight,c(1,1))