calc_EG: Function to calculate the conditional expected genotype...

In Struglab/RVS: Robust Variance Score (RVS) test for association of variants from next generation sequence data with binary traits using external controls

Description

Calculate the conditional expected genotype probability given the observed sequencing data D_{ij} from formula E(P(G_{ij}|D_{ij}))=∑_{g=0}^2 gP(G_{ij}=g|D_{ij}), where P(G_{ij}=g|D_{ij})=P(D_{ij}|G_{ij}=g)*P(G_{ij}=g)/P(D_{ij}). P(D_{ij}|G_{ij}=g) (input M) is from the VCF file or function calc_pobs_ndepth, p(G_{ij}=g) (input p) is output from function calc_EM. All values for P(G_{ij}=g|D_{ij}) are scaled by P(D_{ij}).

Usage

calc_EG(M, p, rdv)

Arguments

M	genotype likelihoods with dimension of number of sample times 3 (double), each column is the genotype likelihood P(D_{ij}|G=AA, Aa\ or\ aa) for one locus. It uses output from get_exp_geno or get_exp_MAF for VCF input.
p	genotype frequencies P(G=AA, Aa\ or\ aa) or p(G=0,1,2\ minor\ allele) for each SNP, it is a vector of length 3 and uses the output from function calc_EM.
rdv	read depth for all samples. Dummy variable in calc_EG, listed here to be consitent with calc_EG_general. It is a vector of integers of one with length equal to the number of samples.

Value

a vector with the same length as rdv, containing conditional expectation probability E(P(G_{ij}|D_{ij})).

Struglab/RVS documentation built on May 9, 2019, 3:11 p.m.