FindMyFriends (version 1.2.2)

gpcGrouping: Guided Pairwise Comparison grouping of genes

Description

This algorithm recursively builds up a pangenome by merging subpangenomes. The recursion follows either a supplied hierarchical clustering or one created using kmer comparison for the full organism. At each step a representative for each gene group is selected randomly as a representative and gets compared to all other representatives. Gene groups are then merged based on the pangenome created for the representatives. Due to the sampling of representatives at each step there is a certain randomness to the algorithm. Results should be fairly stable though, as gene groups are compared multiple times.

Usage

gpcGrouping(object, ...)
"gpcGrouping"(object, lowMem, kmerSize, tree, lowerLimit, pParam, cacheDB, precluster = TRUE, ...)

Arguments

object

A pgVirtual subclass

...

parameters passed on.

lowMem

logical. Should low memory footprint be ensured over computation speed

kmerSize

The size of the kmer's used for the comparison. If two values are given and the 'tree' argument is missing, the second value is used for tree generation. If only one value is given it is recycled.

tree

An optional tree of class dendrogram (or that can be coerced to one) to guide the recursive algorithm. If none is supplied it will be generated by clustering the organisms by their total kmer numbers (summing up for each of their genes).

lowerLimit

A numeric giving the lower bounds of similarity below which it will be set to zero.

pParam

An optional BiocParallelParam object that defines the workers used for parallelisation.

cacheDB

A filehash object or a path to a directory where cached results should be stored. If omitted caching will not be done. Highly recommended for long running instances.

precluster

Logical. Should genes be preclustered using CD-Hit. Defaults to TRUE.

Value

An object of the same class as 'object'.

Methods (by class)

• pgVirtual: gpc grouping for all pgVirtual subclasses

See Also

Other grouping algorithms: cdhitGrouping, graphGrouping, manualGrouping

Examples

testPG <- .loadPgExample()

# Too heavy to include
## Not run: 
# testPG <- gpcGrouping(testPG)
# ## End(Not run)