Convert multi-way concatemer reads to pairwise Hi-C interactions

Transforms multi-way chromatin contact reads (concatemers from Pore-C, Tri-C, Multi-C, etc.) into traditional pairwise Hi-C format. Generates all pairwise combinations of fragments within each read, bins to specified resolution, and creates a GInteractions object compatible with standard Hi-C analysis tools.

Usage

concatemers2Gis(grs, region = NULL, bin_size = 1e+05, read_group = "read_name")

Arguments

grs

GRanges object containing concatemer fragment coordinates. Must include metadata column identifying which fragments belong to the same read (specified by read_group parameter).

region

GRanges object defining genomic region(s) of interest. Only reads overlapping this region are included (default: NULL for genome-wide analysis).

bin_size

Integer. Genomic bin size in base pairs for aggregating contacts. Typical values: 5000-1000000 (default: 100000 = 100kb).

read_group

Character. Name of metadata column in grs that contains read identifiers grouping fragments from the same sequencing read (default: "read_name").

Value

GInteractions object containing:

• Binned pairwise interactions from all concatemers

• count metadata column: number of reads supporting each interaction

• Compatible with gghic(), ChromatinContacts(), and other Hi-C tools

Details

Algorithm

1. Group fragments by read identifier (read_group)

2. For each read with ≥2 fragments, generate all pairwise combinations

3. Bin both anchors of each pair to specified resolution

4. Aggregate identical bin pairs, counting supporting reads

5. Remove self-interactions (same bin pairs)

Multi-way to pairwise conversion

A read contacting N fragments generates N×(N-1)/2 pairwise interactions. For example:

• 2-way contact → 1 pair

• 3-way contact → 3 pairs

• 4-way contact → 6 pairs

This inflation should be considered when comparing multi-way and traditional Hi-C datasets.

Memory considerations

For large datasets, use region parameter to process chromosomes individually and reduce memory usage.

See also

ChromatinContacts(), gghic(), MultiWayContacts()

Examples

if (FALSE) { # \dontrun{
# Load concatemer data
concatemers <- readRDS("concatemers.rds")

# Convert to pairwise at 100kb resolution
gis <- concatemers2Gis(concatemers, bin_size = 100000)

# Visualize as Hi-C map
gghic(gis)

# Focus on specific region at higher resolution
region <- GRanges("chr1:1000000-5000000")
gis_region <- concatemers2Gis(
  concatemers,
  region = region,
  bin_size = 10000
)

# Process by chromosome to manage memory
chr_gis <- lapply(paste0("chr", 1:22), function(chr) {
  region <- GRanges(seqnames = chr,
                    ranges = IRanges(1, 250000000))
  concatemers2Gis(concatemers, region, bin_size = 100000)
})

# Use with ChromatinContacts
gis_all <- do.call(c, chr_gis)
# Then visualize or analyze
} # }