mclustRestricted: mclustRestricted

In scDD: Mixture modeling of single-cell RNA-seq data to identify genes with differential distributions

Description

Function to determine how many normal mixture components are present.

Usage

mclustRestricted(y, restrict = TRUE, min.size)

Arguments

y	Numeric vector of values to fit to a normal mixture model with Mclust.
restrict	Logical indicating whether or not to enforce the restriction on cluster separation based on bimodal index and ratio of largest to smallest variance (see details). If False, then Mclust results as is are returned.
min.size	a positive integer that specifies the minimum size of a cluster (number of cells) for it to be used during the classification step. A clustering with all clusters of size less than min.size is not valid and clusters will be merged if this happens.

Details

Robust to detecting multiple components that are close together by enforcing that the distance between two clusters of appreciable size (at least 4 samples), have sufficiently high bimodal index (cluster mean difference standardized by average standard deviation and multiplied by a balance factor which is one when clusters are perfectly balanced) and not have variances that differ by more than a ratio of 20. Bimodal index threshold is dependent on sample size to ensure consistent performance in power and type I error of detection of multiple components.

Value

List object with (1) vector of cluster membership, (2) cluster means, (3) cluster variances, (4) number of model parameters, (5) sample size, (6) BIC of selected model, and (6) loglikelihood of selected model.

References

Korthauer KD, Chu LF, Newton MA, Li Y, Thomson J, Stewart R, Kendziorski C. A statistical approach for identifying differential distributions in single-cell RNA-seq experiments. Genome Biology. 2016 Oct 25;17(1):222. https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016- 1077-y

scDD documentation built on Nov. 8, 2020, 7:10 p.m.