generate.seeds: Generate seed vectors for the Iterative Signature Algorithm

Description

Generate random input seeds for the ISA.

Usage

generate.seeds (length, count = 100, method = c("uni"), sparsity=2)

Arguments

length

The length of the seeds, should be the number of rows in your input data for row seeds and the number of columns for column seeds.

count

The number of seeds to gnerate.

method

The method for generating the seeds. Currently only "uni" is supported, that picks the 1 elements in each seed uniformly randomly.

sparsity

A numeric scalar, an integer number giving the number of non-zero values in each seed vector. It will be recycled to have the same length as the number of seeds.

Value

A numeric matrix with 0/1 values.

Details

This function can generate a 0/1 matrix whose columns are the seeds of the ISA. The result can be use as the row.seeds (or col.seeds) argument of the isa.iterate function.

References

Bergmann S, Ihmels J, Barkai N: Iterative signature algorithm for the analysis of large-scale gene expression data Phys Rev E Stat Nonlin Soft Matter Phys. 2003 Mar;67(3 Pt 1):031902. Epub 2003 Mar 11. Ihmels J, Friedlander G, Bergmann S, Sarig O, Ziv Y, Barkai N: Revealing modular organization in the yeast transcriptional network Nat Genet. 2002 Aug;31(4):370-7. Epub 2002 Jul 22

Ihmels J, Bergmann S, Barkai N: Defining transcription modules using large-scale gene expression data Bioinformatics 2004 Sep 1;20(13):1993-2003. Epub 2004 Mar 25.

Examples

Run this code

## Just to get always the same result
set.seed(24)

## Create some random seeds with different sparseness
data <- isa.in.silico()
sparsity <- rep( c(1,5,25,125), length=100)
row.seeds <- generate.seeds(length=nrow(data[[1]]), count=100,
                            sparsity=sparsity)

## Do ISA with the seeds
normed.data <- isa.normalize(data[[1]])
isaresult <- isa.iterate(normed.data, thr.row=1, thr.col=1,
                         row.seeds=row.seeds)

## Add the sparsity to the seed data
isaresult$seeddata$sparsity <- sparsity

## Check which ones leed to higher robustness scores
rob <- robustness(normed.data, isaresult$rows, isaresult$columns)
tapply(rob, sparsity, mean)
## About the same

## How many unique modules did we find for the different sparsity
isaresult.unique <- isa.unique(normed.data, isaresult)
tapply(seq_len(ncol(isaresult.unique$rows)),
       isaresult.unique$seeddata$sparsity, length)
## We usually find more modules with sparser seeds

Run the code above in your browser using DataCamp Workspace