Extract sample-level subgraphs from a graph object
Source:R/get_sample_subgraph.R
get_sample_subgraph.RdFor each sample (column in mat), OTUs whose abundance exceeds
min_abundance are treated as present in that sample. The induced
subgraph of graph_obj on those OTUs is returned. Optionally,
multiple selected samples can be combined into a single merged subgraph
via union or intersection of their present OTUs.
Usage
get_sample_subgraph(
graph_obj,
mat,
min_abundance = 0,
select_sample = NULL,
combine = c("union", "intersect")
)Arguments
- graph_obj
A
tbl_graphobject frombuild_graph_from_matorbuild_graph_from_df. Its node table must contain anamecolumn matchingrownames(mat).- mat
Numeric matrix. Rows are OTUs / features (must have rownames matching graph node names); columns are samples (must have colnames as sample IDs).
- min_abundance
Numeric (default = 0). An OTU is considered present in a sample when
mat[OTU, sample] > min_abundance. The appropriate value depends on the scale ofmat: e.g.0for raw or rarefied counts,0.001for relative abundance. The function does not infer data type; the user is responsible for choosing a meaningful threshold for their data.- select_sample
Character vector (default =
NULL). Sample IDs to extract into a single merged subgraph. Must be a subset ofcolnames(mat). WhenNULL, only per-sample subgraphs are returned andsub_graph_selectisNULL.- combine
Character. One of
"union"(default) or"intersect". Controls how OTUs from the selected samples are combined when buildingsub_graph_select:"union": keep OTUs present in any of the selected samples."intersect": keep only OTUs present in all of the selected samples.
Edges are always the induced edges from
graph_objbetween the surviving nodes.
Value
A list with three elements:
sub_graph_all: named list of per-sampletbl_graphsubgraphs. Samples with no present OTUs in the graph are dropped from this list (but are still recorded instat_sample).stat_sample: data frame with columnsSample,Node,Edge,Status. One row per sample inmat, including samples with empty subgraphs.sub_graph_select: a single mergedtbl_graphof the nodes combined fromselect_sampleaccording tocombine, orNULLifselect_sampleisNULLor the combination produces an empty node set. The node table of this graph carries two additional columns:n_present_samples(integer): how many of the selected samples this node appears in.present_in_samples(character): comma-separated sample IDs the node appears in, ordered followingselect_sample.
Details
This function is the sample-wise analogue of get_subgraph,
which splits a graph by its Modularity node attribute. Unlike module
membership (which is a partition), an OTU can belong to multiple samples,
so a combine switch is exposed.
Examples
# \donttest{
data("otu_rare_relative")
data("tax_tab")
obj <- build_graph_from_mat(
mat = otu_rare_relative,
transfrom.method = "none",
r.threshold = 0.7,
p.threshold = 0.05,
method = "WGCNA",
cor.method = "pearson",
proc = "bonferroni",
module.method = "Fast_greedy",
node_annotation = tax_tab,
top_modules = 15,
seed = 1115
)
# All per-sample subgraphs + per-sample stats
res <- get_sample_subgraph(graph_obj = obj, mat = otu_rare_relative)
head(res$stat_sample)
#> Sample Node Edge Status
#> 1 KO1 131 187 OK
#> 2 KO2 141 194 OK
#> 3 KO3 116 404 OK
#> 4 KO4 110 126 OK
#> 5 KO5 85 73 OK
#> 6 KO6 114 146 OK
# Merged subgraph for 3 samples, union of present OTUs
res_u <- get_sample_subgraph(
graph_obj = obj,
mat = otu_rare_relative,
select_sample = colnames(otu_rare_relative)[1:3],
combine = "union"
)
res_u$sub_graph_select
#> # A tbl_graph: 201 nodes and 660 edges
#> #
#> # An undirected simple graph with 36 components
#> #
#> # Node Data: 201 × 16 (active)
#> name modularity modularity2 modularity3 Modularity Degree Strength Kingdom
#> <chr> <fct> <ord> <chr> <ord> <dbl> <dbl> <chr>
#> 1 ASV_12… 6 6 6 6 29 28.3 Bacter…
#> 2 ASV_14… 6 6 6 6 28 27.4 Bacter…
#> 3 ASV_649 6 6 6 6 27 26.5 Bacter…
#> 4 ASV_705 6 6 6 6 27 26.5 Bacter…
#> 5 ASV_913 6 6 6 6 27 26.2 Bacter…
#> 6 ASV_13… 6 6 6 6 27 26.5 Bacter…
#> 7 ASV_14… 6 6 6 6 27 26.5 Bacter…
#> 8 ASV_17… 6 6 6 6 27 26.3 Bacter…
#> 9 ASV_24… 6 6 6 6 27 26.5 Bacter…
#> 10 ASV_25… 6 6 6 6 27 26.4 Bacter…
#> # ℹ 191 more rows
#> # ℹ 8 more variables: Phylum <chr>, Class <chr>, Order <chr>, Family <chr>,
#> # Genus <chr>, Species <chr>, n_present_samples <int>,
#> # present_in_samples <chr>
#> #
#> # Edge Data: 660 × 5
#> from to weight correlation corr_direction
#> <int> <int> <dbl> <dbl> <chr>
#> 1 151 152 0.959 0.959 Positive
#> 2 151 153 0.941 0.941 Positive
#> 3 151 154 0.940 0.940 Positive
#> # ℹ 657 more rows
# Same selection but intersection (core OTUs across the 3 samples)
res_i <- get_sample_subgraph(
graph_obj = obj,
mat = otu_rare_relative,
select_sample = colnames(otu_rare_relative)[1:3],
combine = "intersect"
)
res_i$sub_graph_select
#> # A tbl_graph: 63 nodes and 43 edges
#> #
#> # An undirected simple graph with 30 components
#> #
#> # Node Data: 63 × 16 (active)
#> name modularity modularity2 modularity3 Modularity Degree Strength Kingdom
#> <chr> <fct> <ord> <chr> <ord> <dbl> <dbl> <chr>
#> 1 ASV_12… 6 6 6 6 29 28.3 Bacter…
#> 2 ASV_913 6 6 6 6 27 26.2 Bacter…
#> 3 ASV_767 6 6 6 6 22 21.1 Bacter…
#> 4 ASV_322 6 6 6 6 19 18.1 Bacter…
#> 5 ASV_18… 6 6 6 6 19 18.2 Bacter…
#> 6 ASV_277 6 6 6 6 4 3.76 Bacter…
#> 7 ASV_244 6 6 6 6 3 2.87 Bacter…
#> 8 ASV_367 6 6 6 6 3 2.83 Bacter…
#> 9 ASV_132 6 6 6 6 1 0.941 Bacter…
#> 10 ASV_927 10 10 10 10 16 15.6 Bacter…
#> # ℹ 53 more rows
#> # ℹ 8 more variables: Phylum <chr>, Class <chr>, Order <chr>, Family <chr>,
#> # Genus <chr>, Species <chr>, n_present_samples <int>,
#> # present_in_samples <chr>
#> #
#> # Edge Data: 43 × 5
#> from to weight correlation corr_direction
#> <int> <int> <dbl> <dbl> <chr>
#> 1 1 2 0.986 0.986 Positive
#> 2 1 3 0.965 0.965 Positive
#> 3 2 3 0.967 0.967 Positive
#> # ℹ 40 more rows
# }