iProFun.prob: Calculating posterior probabilities of association patterns

In songxiaoyu/iProFun: An integrative analysis tool to screen for Proteogenomic Functional traits perturbed by DNA alterations

Calculating posterior probabilities of association patterns

Description

This is the core function of iProFun, and it calculates the posterior probabilities of association patterns from regression summaries from multiple data types. It allows some data types from regression to opt out from calculating posterior probabilities, which is suitable for the data types with very few genes (e.g. somatic mutation) that borrowing information across variables is not reliable.

Usage

iProFun.prob(Reg.Sum, NoProbXIndex = NULL, pi1 = 0.05)

Arguments

Reg.Sum	Linear regression analysis summaries formatted in multi.omic.reg.summary.
NoProbXIndex	NoProbXIndex allows users to provide the index for the predictor data type(s) that are not considered for calculating posterior probabilities of association patterns. NoProbXIndex = NULL (default): all data types in predictors will be considered. Any 0 < NoProbXIndex <= length of xList: indicates that the posterior probabilities of association patterns for the corresponding data type(s) are not calculated.
pi1	pi1 is the pre-specified prior proportion of non-null statistics. It cane be a number in (0, 1) or a vector of numbers with length of ylist.

Value

A list with the same length as xlist. Nested within each list, it contains

NoComputation:	No of variables considered for this predictor with all outcomes
Config:	Corresponding association patterns. Total number 2^J for J outcome data types
Config.miss:	Corresponding association patterns for variables (e.g. genes) that are missing in some outcome data type(s)
PostProb:	Posterior probability for each predictor on each association pattern
PostProb.miss:	Posterior probability for each predictor on each available association pattern for variables (e.g. genes) that are missing in some outcome data type(s)
xName.miss:	Predictor name for variables (e.g. genes) missing in some data type(s)
colocProb:	Averaged posterior probability for predictor on each association pattern (missing data excluded)
Tstat_L:	T statistics for each predictor on each outcome
D0:	Estimated density under the null for predictor on each outcome
D1:	Estimated density under the alternative for predictor on each outcome

songxiaoyu/iProFun documentation built on Dec. 8, 2022, 3:54 p.m.