RchyOptimyx (version 2.12.0)

RchyOptimyx: Gating Hierarchy Optimization for Flow Cytometry

Description

RchyOptimyx is a computational tool which determines the minimal sets of markers that can identify a target population within a given purity standard. This can enable panel redesign using smaller subsets of markers, with consequent savings on reagents, potential for adaptation to older instruments, or the addition of other markers of interest to the panel design.

Usage

"RchyOptimyx"(pheno.codes, phenotypeScores, startPhenotype, pathCount=1, trimPaths=FALSE, trim.tolerance=0, trim.level=0)

Arguments

pheno.codes

A vector of strings of length N (the number of markers) for each phenotype measured. For every phenotype, the character corresponding to a given marker can be 0, 1, 2, etc for neutral, negative, positive, bright, etc. See the provided vigenette (inst/doc/RchyOptimyx.pdf) for more details and examples.

phenotypeScores

A vector of the scores assigned to every phenotype. The optimization algorithm will try to maximize this score. See the provided vigenette (inst/doc/RchyOptimyx.pdf) for more details and examples.

startPhenotype

The terminal phenotype which includes every marker that must be considered in the analysis. This variable is a vector of the length of the number of markers. Every element of the vector can be 0, 1, or 2 for negative, neutral, or positive respectively (see details).

pathCount

The number of paths that must be generated.

trimPaths

A boolean indicating wether the paths should be trimmed or not. If set to TRUE, the paths will be terminated as soon as adding trim.tolerance extra number of edges to the path doesn't result in an improvement in the score.

trim.tolerance

An integer indicating the number of levels before the path is trimed if the score is decreasing by adding extra edges.

trim.level

An integer indicating from which level the tree is going to be trimed. This parameter has no effect if set to zero.

Value

OptH

An OptimizedHierarchy object

Details

Every marker can be positive, negative, or neutral. Neutral markers are not included in the measurement of the respective phenotype. A path is trimed in one of the following conditions: 1. If the trimPaths parameter is TRUE, the path is trimed as soon as it decreases for equal or more levels that trim.trolerance parameter. If the path is decreasing while reaching the trim.level, the path is trimed from the point that it started to decrease. 2. If the path reaches the trim.level.

References

D. Eppstein, Finding the k shortest paths, SIAM J. Comput., 1998 - Citeseer.

Nima Aghaeepour, Pratip K. Chattopadhyay, Anuradha Ganesan, Kieran O'Neill, Habil Zare, Holger H. Hoos, Mario Roederer, and Ryan R. Brinkman. Early Immunologic Correlates of HIV Protection can be Identified from Computational Analysis of Complex Multivariate T-cell Flow Cytometry Assays. Bioinformatics, Feb. 2012.

Examples

data(HIVData)

res<-RchyOptimyx(PhenoCodes, -log10(LogRankPvals), '2000102001', 5,FALSE) 
plot(res,phenotypeCodes=PhenoCodes,phenotypeScores=-log10(LogRankPvals),marker.names=marker.names)