Skip to content

Estimation and sampling for multi-dimensional Hurdle models on a Normal density with applications to single-cell co-expression

Notifications You must be signed in to change notification settings

amcdavid/HurdleNormal

Repository files navigation

HurdleNormal

Estimation and sampling for multi-dimensional Hurdle models on a Normal density with applications to single-cell co-expression

A Hurdle model is a modification of a distribution at zero. This package provides routines to estimate and sample from multivariate Hurdle models on a Normal density. These are distributions that are conditionally Normal, but with singularities along the coordinate axes, so generalize a univariate zero-inflated distribution.
The main functionality is to estimate graphical models using group-lasso penalized neighborhood selection.

Install

BiocManager::install('amcdavid/HurdleNormal')

or

BiocManager::install('amcdavid/HurdleNormal', dependencies = TRUE)

if you want to compile the vignette or run examples from the paper.

Vignette

A vignette with an example is located here or with vignette('singlecell-networks').

Citation

Graphical Models for Zero-Inflated Single Cell Gene Expression

About

Estimation and sampling for multi-dimensional Hurdle models on a Normal density with applications to single-cell co-expression

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published