Marginal PP for models sharing information between traits

Marginal PP for models sharing information between traits

marginalpp(
  STR,
  PP,
  mbeta,
  covY,
  SSy,
  Sxy,
  kappa,
  N,
  Nqq,
  nsnps,
  Mx,
  xcovo,
  Nq3,
  Nq4,
  fastapprox,
  NCORES
)

Arguments

STR

list of models for traits 1, 2, ..., n, each given in the form of a character vector, with entries "snp1%snp2%snp3". The null model is given by "1" OR "0". It is assumed that all elements of ABF, PP and pr below follow this same order.

PP

list of posterir probabilities for the models in M

mbeta

list of joint beta estimates for each trait

covY

trait covariance matrix

SSy

matrix of trait cross-products

Sxy

matrix with each column being the cross-product between SNPs and a trait

kappa

single value or vector of values to consider for the sharing scale parameter. the value of kappa=1 must be included, and if not will be prepended.

N

number of individiduals with measurements for all traits

Nqq

matrix in which Nqq[i,j] = number of individuals measured in both trait i and trait j

nsnps

number of snps in region

Mx

vector of SNP means

xcovo

SNP covariance matrix

Nq3

vector of counts of number of individuals with three traits measured; all triples considered; NULL if M < 4

Nq4

vector of counts of number of individuals with four traits measured; all quadruples considered; NULL if M < 5

fastapprox

logical that is TRUE when fast approximation is used that does not include unequal sample size adjustments; default is FALSE

NCORES

number of cores for parallel computing; recommend NCORES=M, but if on Windows, use NCORES=1

Value

list of: - single.pp: list of pp for each model in STR[[i]] for trait i - shared.pp: list of pp for each model in STR[[i]] for trait i