GenomicOZone: Delineate outstanding genomic zones
In GenomicOZone: Delineate outstanding genomic zones of differential gene activity

Description Usage Arguments Details Value References See Also Examples

View source: R/MD_main.R

Delineate outstanding genomic zones along chromosomes such that genes within an outstanding zone have consistent activity patterns that are different across samples.

1	GenomicOZone(GOZ.ds)

GOZ.ds

an object created by function GOZDataSet.

This is the most important function of the package. It integrates genome annotation, gene activity matrix preprocessing, chromosome clustering, and differential zone analysis.

Genome annotation can be specified either by the user, or obtained from the R package biomaRt \insertCitesmedley2015biomartGenomicOZone to access ensembl annotation databases \insertCitezerbino2017ensemblGenomicOZone.

The function calls the weighted univariate clustering method \insertCitewang2011ckmeansGenomicOZone implemented in the Ckmeans.1d.dp package. If ks is specified, a fixed number of zones at each chromosome will be delineated. If ks is NULL, an optimal number of clusters at each chromosome will be determined by Bayesian information criterion.

The function also conducts differential zone analysis by using one-way ANOVA \insertCitechambers1992statisticalGenomicOZone based on gene ranks. Given p-value cutoff alpha and effect size threshold min.effect.size, outstanding genomic zones will be selected.

Advanced differential zone analysis such as generalized linear modeling can be performed on the zone activity matrix using third-party software. The zone activity matrix can be generated by auxiliary functions.

an object which is the input object attached with intermediate and final results. Results can be accessed by calling several functions in extract_outputs. The results can be visualized by calling the functions in generate_plots.

\insertAllCited

See GOZDataSet for how to create the input list. See extract_outputs and generate_plots for how to access the results and generate visualizations.

  # Create an example of GOZ.ds
  data <- matrix(c(1,5,2,6,5,1,6,2), ncol = 2, byrow = TRUE)
  rownames(data) <- paste("Gene", 1:4, sep='')
  colnames(data) <- paste("Sample", c(1:2), sep='')

  colData <- data.frame(Sample_name = paste("Sample", c(1:2), sep=''),
                        Condition = c("Cancer", "Normal"))

  design <- ~ Condition

  rowData.GRanges <- GRanges(seqnames = Rle(rep("chr1", 4)),
                             ranges = IRanges(start = c(1,2,3,4), end = c(5,6,7,8)))
  names(rowData.GRanges) <- paste("Gene", 1:4, sep='')

  ks <- c(2)
  names(ks) <- "chr1"

  GOZ.ds <- GOZDataSet(data, colData, design,
                       rowData.GRanges = rowData.GRanges,
                       ks = ks)
  ####

  # Run the zoing process
  GOZ.ds <- GenomicOZone(GOZ.ds)
  ####