Construct a DMSA reference bundle

Usage

dmsa_reference(
  systems,
  polarity = NULL,
  anchors = NULL,
  anchor_method = c("curated", "graph_sink", "user", "none"),
  name = "unnamed",
  version = NA_character_,
  notes = NULL
)

Arguments

systems: data.frame with columns gene, system_id, system, and optionally module_id and module. One row per gene-within-system (a gene may appear in several systems).
polarity: data.frame with columns gene, system_id, w_g. w_g may be -1/0/+1 or continuous in [-1, 1]. Optional columns evidence, role, confidence, source are carried through.
anchors: Optional data.frame with system_id, gene - the genes that define the direction of "more activation" for each system.
anchor_method: Character, how the anchors were obtained. One of "curated", "graph_sink", "user", or "none" (systems then carry no sign and only gene-level analysis is valid).
name, version, notes: Provenance, printed and stored.

Value

An object of class dmsa_reference.

Examples

## One row per gene-within-system, a sign per gene, and the anchor genes that
## fix which way "more activation" points for each system.
sys <- data.frame(gene = c("CRH", "POMC", "NR3C1", "OXT", "OXTR"),
                  system_id = c("1", "1", "1", "2", "2"),
                  system = c(rep("HPA axis", 3), rep("Oxytocin", 2)))
pol <- data.frame(gene = sys$gene, system_id = sys$system_id,
                  w_g = c(1, 1, -1, 1, 0.6))
anc <- data.frame(system_id = c("1", "2"), gene = c("CRH", "OXT"))
dmsa_reference(sys, pol, anc, anchor_method = "curated", name = "toy",
               version = "0.1")
#> dmsa reference: toy (0.1)
#>   5 genes across 2 systems  (no module layer)
#>   polarity: 5 gene-system pairs, continuous (4 activating, 1 braking, 0 off-axis)
#>   anchors: curated (2 genes)