seqlm: Sequential lm
In raivokolde/seqlm: A method to find differentially methylated regions.

Description Usage Arguments Details Value Author(s) Examples

Segments genome based on given linear models and and calculates the significance of regions

seqlm

1 2	seqlm(values, genome_information, annotation, max_block_length = 50, max_dist = 1000)

`values`	a matrix where columns are samples and rows correspond to the sites
`genome_information`	`GRanges` object giving the positions of the probes, names should correspond to rownames of values. `elementData` of this object is used to annotate the regions
`annotation`	vector describing the samples. If discrete then has to have exactly 2 levels.
`max_block_length`	maximal length of the block we are searching. This is used to speed up computation
`max_dist`	maximal genomic distance between the sites to be considered the same region

The analysis can be time consuming if the whole genome is analysed at once. If the computer has multicore capabilities it is easy to parallelize the calculations. We use the foreachframework by Revolution Computing for parallelization. To enable the parallelization one has to register the parallel backend before and this will be used by seqlm.

A list containing the input data, parameters and the segmentation.

Kaspar Martens <kmartens@ut.ee> Raivo Kolde <rkolde@gmail.com>

data(artificial)
seqlm(artificial$values, artificial$genome_information, artificial$annotation1)

## Not run: 
data(tissue_small)

# Find regions
segments = seqlm(tissue_small$values, tissue_small$genome_information, tissue_small$annotation)

# The calculation can be parallelized by registering a parallel processing backend
library(doParallel)
registerDoParallel(cores = 2)
segments = seqlm(values = tissue_small$values, genome_information = tissue_small$genome_information, annotation =  tissue_small$annotation)

# To visualise the results it is possible to plot the most imortant sites and generate a HTML report
temp = tempdir()
seqlmreport(segments[1:10], tissue_small$values, tissue_small$genome_information, tissue_small$annotation, dir = temp)

# To see the results open the index.html file generated into the directory temp

## End(Not run)