Build a network from odds ratio data.

Construct a tidygraph network from pairwise odds ratio or ROR data. Each node is a position (or isodecoder) and each edge represents a significant co-occurrence relationship. Requires the tidygraph and igraph packages.

Usage

build_or_network(data, value_col = "ror", min_weight = 0)

Arguments

data

A tibble with columns pos1, pos2, and a value column specified by value_col. Optionally includes p_adj.

value_col

Column name (string) for edge values. Default "ror".

min_weight

Minimum absolute value of value_col for an edge to be included. Default 0.

Value

A tbl_graph object with node metrics degree and betweenness, or NULL if no edges remain after filtering.

Examples

# \donttest{
df <- tibble::tibble(
  pos1 = c("20", "34", "20"),
  pos2 = c("34", "58", "58"),
  ror = c(1.5, -0.8, 0.3)
)
build_or_network(df)
#> # A tbl_graph: 3 nodes and 3 edges
#> #
#> # An undirected simple graph with 1 component
#> #
#> # Node Data: 3 × 3 (active)
#>   position degree betweenness
#>   <chr>     <dbl>       <dbl>
#> 1 20            2           0
#> 2 34            2           0
#> 3 58            2           0
#> #
#> # Edge Data: 3 × 4
#>    from    to edge_type    weight
#>   <int> <int> <chr>         <dbl>
#> 1     1     2 co-occurring    1.5
#> 2     2     3 exclusive       0.8
#> 3     1     3 co-occurring    0.3
# }