[Bioc-devel] flowMerge 0.3.5 now available

[Greg Finak]

Hello, everyone.
The computational biology group at the IRCM is announcing the  
availability of flowMerge, a new BioConductor package for automated  
gating of flow cytometry data.
The package uses the flowClust framework to fit multivariate-t  
mixtures with Box-Cox transformation to flow data.
Frequently, with complex cell populations, multiple components are  
required to model a single population. flowMerge uses an entropy  
criterion to merge overlapping mixture components from the best  
fitting (max BIC) model to arrive at a better estimate of the number  
of distinct cell populations.  Because the parameters of merged  
clusters (mean, covariance, degrees of freedom) can be derived from  
their merged component clusters, these complex cell populations can be  
represented by a single mixture component, and retain the good model- 
fitting properties of the max BIC model selection.  Functions are  
provided to allow calculations in parallel using the snow package and  
the package integrates into the existing flowCore and flowClust  
frameworks.
NOTE: You will need flowClust 2.0 to use this package.

Cheers,

Greg Finak, PhD
Post-Doctoral Research Associate
Computational Biology Unit
Institut de Recherches Cliniques de Montréal
greg.finak at ircm.qc.ca