GPATreeStage2: Implement the Stage 2 of the GPA-Tree approach.

In asthakhatiwada/GPATree: A package to implement the GPA-Tree method

Description

This function will implement the Stage 2 of the GPA-Tree approach.

Usage

GPATreeStage2(gwasPval, annMat, alphaStage1, initPi, cpTry)

Arguments

gwasPval	A matrix of M X 1 dimension, where M is the number of SNPs. The first column is labeled 'SNPid' and contains the SNPid. The second column contains the GWAS association p-values and is called 'P1'. The values in P1 must be between 0 and 1.
annMat	A matrix of binary annotations, where row and column correspond to SNPs and annotations, respectively.
alphaStage1	alpha estimated in stage 1 of the GPA-Tree approach.
initPi	pi estimated in stage 1 of the GPA-Tree approach.
cpTry	Complexity parameter (cp) value to be used to build annotation decision tree. cpTry can be between 0 and 1 or NULL. Default is 0.001. When cpTry is NULL, GPATree will select the optimal cp to be used.

Value

This function returns a List including:

• num_iter_stage2: number of iterations taken for Stage 2 of GPA-Tree approach to converge.

• lic_list_stage2: list of incomplete log-likelihood for all iterations in Stage 2 of GPA-Tree approach.

• lc_list_stage2: list of complete log-likelihood for all iterations in Stage 2 of GPA-Tree approach.

• zi_stage2: probability of being a null and non-null SNP.

• Zmarg: marginal posterior probability of being a non-null SNP for the phenotype.

• pi_stage2: updated prior.

• var_in_tree: annotations included in the tree.

• pruned_tree: GPA-Tree model fit.

Author(s)

Aastha Khatiwada

asthakhatiwada/GPATree documentation built on March 8, 2021, 5:29 a.m.