Normalize intervals to fixed or variable sizes

This function normalizes intervals by computing their centers and then expanding them to fixed or variable sizes, while ensuring they don't cross chromosome boundaries.

Usage

gintervals.normalize(intervals = NULL, size = NULL, intervals.set.out = NULL)

Arguments

intervals

intervals set

size

target size(s) for normalized intervals, before the clipping at chromosome ends. Can be either:

• A single positive integer: all intervals normalized to this size

• A numeric vector: each interval normalized to its corresponding size. Vector length must exactly match the number of intervals, OR a single interval can be provided with multiple sizes to create multiple output intervals (one-to-many expansion).

intervals.set.out

intervals set name where the function result is saved. If NULL, the result is returned to the user.

Value

Normalized intervals set with fixed or variable sizes, or NULL if result is saved to intervals.set.out

Details

A window that would cross a chromosome end is clipped at that end, so near the start or the end of a chromosome the returned interval is shorter than 'size' (no warning is given). Filter on end - start if the windows must all have the full size.

See also

gintervals.force_range

Examples

gdb.init_examples()

# Single size (all intervals normalized to 500bp)
intervs <- gintervals(1, c(1000, 5000), c(2000, 6000))
gintervals.normalize(intervs, 500)
#>   chrom start  end
#> 1  chr1  1250 1750
#> 2  chr1  5250 5750

# Vector of sizes (each interval gets its own size)
intervs <- gintervals(1, c(1000, 3000, 5000), c(2000, 4000, 6000))
gintervals.normalize(intervs, c(500, 1000, 750))
#>   chrom start  end
#> 1  chr1  1250 1750
#> 2  chr1  3000 4000
#> 3  chr1  5125 5875

# One-to-many: single interval with multiple sizes
interv <- gintervals(1, 1000, 2000)
gintervals.normalize(interv, c(500, 1000, 1500))
#>   chrom start  end
#> 1  chr1  1250 1750
#> 2  chr1  1000 2000
#> 3  chr1   750 2250