R: Fitting a beta-mixture model to association scores
Mulder2012.BMfitting
R Documentation
Fitting a beta-mixture model to association scores
Description
The function fits a three-beta mixture model to functional association scores
computed from rich phenotyping screens in Mulder K et al., 2012 and Wang X et
al., 2012.
a numeric matrix of z-scores with rows and columns corresponding to samples and
genes, respectively.
model
a character value specifying whether the original (if global) or extended
(if stratified) model is used.
metric
a character value specifying the metric to compute similarity scores. Currently,
'cosine' and 'correlation' are supported (see assoScore for more details).
nPerm
a positive numeric or integer value specifying the number of permutations.
partition
a numeric of gene partition labels (e.g. c(rep(1, 100), rep(2, 20)) is a valid
vector of partition labels for a vector of associations falling into two
categories of interaction types 1 and 2).
Details
The two functions implement beta-mixture modelling in the application to epidermal
stem cells.
Both functions first initiate an object of class BetaMixture with
inputted pheno, model and metric (and partition for the
stratified model).
The ‘x’ component of the beta-mixture model is then fitted to permuted data for
nPerm times, and the estimated shape parameters of beta distributions are
summarized by taking the median values.
After fixing the shape parameters of the ‘x’ component, the function invokes
fitBM to fit the global or stratified beta-mixture model.
Value
an object of class BetaMixture with beta-mixture modelling
results stored at slot result
Mulder KW, Wang X, Escriu C, et al. (2012). Diverse epigenetic strategies
interact to control epidermal differentiation. Nature Cell Biology.
doi:10.1038/ncb2520.
Wang X, Castro MA, Mulder KW and Markowetz F (2012), Posterior association
networks and enriched functional gene modules inferred from rich phenotypic
perturbation screens, PLoS Computational Biology, doi:10.1371/journal.pcbi.1002566.