Skip to contents

Run the official monocle3-compatible Cicero workflow on the selected ATAC counts and an explicitly selected cell embedding. No Pearson fallback is substituted if Cicero or its model fitting is unavailable.

Usage

RunATACCoaccessibility(
  object,
  genome.sizes,
  genome.id,
  assay = NULL,
  reduction = "lsi",
  dims = 2:30,
  k = 50L,
  window = 500000L,
  sample.num = 100L,
  seed = 1L,
  result.name = "scop_coaccessibility",
  verbose = TRUE
)

Arguments

object

Seurat object with a ChromatinAssay.

genome.sizes

Named positive integer chromosome lengths for the verified reference, covering all source peaks. No reference is downloaded.

genome.id

Reference label matching the selected assay.

assay

ATAC assay; defaults to active assay.

reduction

Existing reduction whose cells cover every assay cell.

dims

Explicit dimensions used for nearest-neighbor aggregation.

k

Number of cells per aggregate; at least 2 and less than cell count.

window

Genomic window length, greater than 250000 because the run_cicero distance constraint is fixed at 250000.

sample.num

Number of windows sampled for distance estimation.

seed

Local random seed. Caller RNG state is restored.

result.name

Unused assay Misc key for the result.

verbose

Print progress messages.

Value

Input Seurat object with a result list in assay Misc(result.name), containing connections, aggregate membership, parameters and source identifiers. Links, source counts, metadata and default assay are unchanged. Receipt identifiers are calculation-time snapshots.

Details

Requires the official Cicero monocle3 branch and monocle3. See the official installation instructions. The installed Cicero package must declare a monocle3 dependency. Binary accessibility is aggregated by Cicero before size-factor normalization and distance-penalized graphical-model fitting. Model fitting failures (including inadequate genomic coverage) are preserved. Peak IDs and cell identities are not renamed or silently dropped. Returned NA scores are unresolved/discordant estimates, not zero. Associations are not experimentally verified physical interactions or causal regulation.